FiRe 2019 in Review, Part II:
About this year's "Future in Review in Review" double issue:
Those SNS members who are repeat FiRe attendees may have noticed that this year you weren't surrounded by the tap-tap-tap of four-plus full-time bloggers. We decided to try a different model for 2019: while encouraging our FiReFellows to spend more time with FiReStarter companies and other attendees, we would focus on capturing just a few bullet points for as many sessions as could be covered, carried off by just two valiant volunteers: Dakota Carter in the lead, with Evan Anderson assisting and following up after the event.
In true FiRe style, it turns out that a request for "a few bullet points" is subject to interpretation - and ambition. As you'll see in the first of this two-part review, reports range from brief prcis to full-length articles to mini-transcriptions to a dozen or so bullet points.
Given these disparate styles, in the end we thought that rather than squeeze them into a one-size-fits-all model, it would make more sense to let them run (with slight edits and embellishments) in their individual styles. That also seemed so ... FiRe.
We hope you enjoy the FiRe team's efforts, founded on passion, appreciation, curiosity, and some revisits to get the facts straight.
Thanks yet again to photographer-sorcerer Kris Krg, the talent behind all of the images here (unless otherwise noted), as well as all those in the FiRe galleries since 2015.
Please refer any comments or corrections to me at firstname.lastname@example.org.
We hope you enjoy entering or re-entering the deep world of FiRe, that you learn something you hadn't previously known you were curious about, and that we see you in person at FiRe 2020, Oct. 6-9, in La Jolla. - Sally Anderson, Editor-in-Chief, SNS, FiRe, and FiReFilms
China's Economic Strategy
Kim Stanley Robinson, Science Fiction Writer; and
Moderated by Kimberly Dozier, Global Affairs Analyst, CNN
by Evan Anderson with Sally Anderson
Kimberly Dozier opened this discussion by posing separate questions to guests Mark Anderson, FiRe chair, and science-fiction writer Kim Stanley Robinson. To Anderson, she noted his frequent discussion of China's economic model in the pages of the SNS Global Report and asked whether he sees our situation with China today as total war - "'total' being Sun Tzu's definition that you can win without fighting on the battlefield, politically, economically, through the gray zones, versus absolute war, in the sense of the Clausewitz term, meaning you can't win an absolute war, a la World War II, when the Nazis tried to declare war on the entire world. At what point do the Chinese hit the tipping point, and how do we navigate that?"
Anderson said that sounded somewhat like the Powell doctrine.
Agreeing, Dozier said she'd then be asking Robinson: "How can we harness the good things in Chinese culture - their love of the outdoors, which we don't see very much in government policy right now, things like the concept of Zen, their pride - wanting to be great creators, not destroyers, to turn this around?"
L-R: Kimberly Dozier, Mark Anderson, and Kim Stanley Robinson
First, some of Anderson's responses:
What we missed entirely in 2000 was the switch to the infomercantilist model in China: "We were thinking of them as communists, economically, and they weren't, then. So we missed it for a long time. Even now, we're missing most of it."
Basic tenet of infomercantilism: If I wanted to design a disease for a human being, I could be pretty clever about it and go after the heart, the lungs, and the brain. That is the infomerc model in the face of capitalism.
Infomercantilism takes all the pieces of capitalism that are useful to the dictators in charge of China and leaves out the honesty, the open markets, and the freedoms that usually come with it. They maintain a one-sided approach, and they turn our companies into parking lots. "From a systems perspective, when we're all gone, or working for them, innovation will be stifled, and we enter a very dark time."
Dozier: Is there something in Chinese culture that can bring some reason to this process?
Anderson:[The Chinese government is] terrified of their own people, so revolution is their number one problem. Everything that Xi Jinping thinks about all day long is in the context of "Don't let them rise up." Every decision. That's the one thing they can do wrong. He knows it, the Standing Committee knows it .... That's their prime motivator in all the decisions they make.
Kim Stanley Robinson:I would say that's right, but one of the prime methods by which you keep the populace happy is to satisfy ... rising expectations. The poverty reduction program in rural parts of China is very real ... a good thing; so much of China is still stupendously poor.
That's one of Xi's strategies to keep the citizens happy, Robinson said. Historically speaking, revolution doesn't come in times of desperation, when people are just trying to make it through the day, but at times when expectations are rising. The Chinese people are "caught in multiple contradictory traps, many of them not of their own making."
The two agreed that although the West had beaten down China during the Century of Humiliation, America itself had been a friend through that time. Robinson added that it was a good relationship even before the revolution, and doesn't think China dislikes America by comparison with its European conquerors in the 19th century.
Dozier: We're trying to fight them off in courts and in trade. Right strategy?
Anderson:Not exactly, but close.... It's an economic attack. So it is a Sun Tzu approach: it is asymmetric, it's very interesting, they use everything except bullets... and it works.
Dozier asked if he meant the Chinese break partnerships, to which Anderson said, "No - they just make them weaker and weaker and weaker and weaker, and then they go away. The only response to an economic attack that works, it turns out, is an economic response." Shame does not work. He understands that it's in China's interest to make its people richer every day; he just wishes it weren't "with our money."
Anderson: It is a world war, already under way, and people are just waking up to it, and that's exactly what Sun Tzu would've hoped. You don't even know you're at war, before they win. And not a shot was ever fired.
It's critically important to have an economic alliance with European and other nations right now, Anderson said, to put more economic pressure on "the guys who are destroying the global economy." Blanket tariffs were a mistake, but it was brilliant to do tariffs.
Dozier suggested that to apply economic pressure "means you've got to acknowledge the attack." Anderson agreed. Dozier added: "Which is tough when you've got China on the UN Security Council at most of the top trade organizations as well." Anderson again agreed, saying that one of the main needs China has with its "fake cash," aside from giving it to its people, is to buy off other countries. First was Britain: "It was really kind of crushing to see our pal Britain become a vassal of China."
Dozier asked Robinson if we understand that we're in a war. Are we psychologically prepared? Are the Chinese?
Robinson:I don't think "war" is the right word for this. For the Chinese, it's kind of a survival game.... We're mutual customers, in that they provided the working class and cheap labor, and all the Western companies went for it, and industry moved there, and a lot of dollars went to China in a perfectly free-trade transaction.
He does agree with Anderson about China's employment of an infomerc model, but doesn't feel the leadership is committed to that. They're improvisers, Robinson said, who might shift if a different model came along that they liked better: "Deng called it 'crossing the river by feeling the stones.' There's no ideological commitment," and no relation to socialism as we tend to think of it. "With Chinese characteristics is the crucial phrase when the Chinese describe themselves."
Robinson asked Anderson if China should've been allowed to join the WTO, and agreed with him when Anderson said No. Where they disagree is that Robinson thinks China will stick to the letter of treaties, but he conceded that they didn't do so with the WTO: "They will push where they can push."
Shalev Lifshitz, AI Researcher and Entrepreneur, University of Waterloo, SickKids Hospital, and St. Joseph's Health Centre;
Moderated by John Wells, Producer, Cool Science Radio, KPCW (NPR Affiliate)
L-R: John Wells, Shalev Lifshitz, Caitlin Cameron, and John Mattison
Shalev Lifshitz: The main question that everyone is asking is, "Is AI good for humanity?" We need to look at its impact; it's being used in a lot of good ways. If we more closely mimic the human brain with neural networks, it should be stronger. But there are many ethical concerns.
Caitlin Cameron:[Today] AI is making great strides to do good. This device we [OtoNexus] currently have is a technology for diagnosing middle-ear infections in children. In the future, when we roll this out, we'll have millions of datapoints in the cloud and can start to do analysis and find the nuances in those datapoints.... I truly believe that AI is going to dramatically improve healthcare.
John Mattison:The term "the singularity" has served its purpose about thinking how humans and machines interact, but I prefer the term "dyadarity" to really reflect the dyad of the human and the machine and the collaboration involved in that.
Lifshitz said that we can make AI more like the brain. "A main topic is the danger of AI, and what could go wrong. I think the main thing about AI is that it's an amplification tool. If you use it for good purposes, you amplify really good things. If you use it for bad purposes, you amplify
bad things. Now, the scary thing is, let's say we reduce all the negative aspects or we make sure people don't use AI for bad purposes. Even if a good person makes a mistake, that mistake is now amplified. So as we create more intelligent technology, our mistakes are amplified and have a deeper effect.
John Wells: How do we align our goals with our smart machines?
Mattison: Goals and values.
Lifshitz: The values could be the way to align the goals, but there are three problems with that. The first is a technical problem: we don't know how to encode values into a machine, values that can't be broken. The second one is a political problem: which values do we choose? Because as we've clearly seen yesterday and today, as cultures change, values change. So is one set of values better than another? How do we choose - should we collaborate?
And the third one is: How do you define a value? Let's say your value is "Don't harm a human" - well, do you have to now define every single possible way to harm a human? Isn't that limiting? We know machines always find pathways through the rules we give them.
Mattison: The cycle time is so fast with AI, but the regulatory response is so slow. There may be room for a consumer-based oversight that looks something like the environmental impact statements of the past... [T]hink about a human impact statement, and if we establish a standard of the proprietors of these machines that they have ... publicly available their human impact statement, where they can articulate what they think are the risks, and have some transparency into how a particular vendor of AI services is thinking about what risks are possible and how they intend to manage against them.
Cameron: I'm not sure how we get China to do that, though.
Lifshitz: I think one of the biggest dangers from AI is not from AI, it's from humans.
The panel further discussed ethical difficulties and challenges in coming to consensus globally. A main note was that healthcare might be a great place to collaborate, because it matters to everyone.
As the panel adjourned, Cameron was asked to stay onstage, where FiRe chair Mark Anderson presented OtoNexus Medical Technologies with the FiRe 2019 Company of the Year award.
Anand S. Rao, Partner, Advisory Services, and Global AI Lead, PwC;
Moderated by Mark Anderson, Chair, Future in Review
This conversation was a grounded reflection of what AI is currently capable of, and more important, what it is not.
L-R: Mark Anderson, Anand S. Rao, Andrew Fano, and Paul Maher
Anand S. Rao, who has worked in the AI field since the 1980s, said there is a long history of overpromising and underperforming. While he hopes we won't experience another "AI winter," he posits that we're likely experiencing an overpromise stage.
Describing AI as "any system that can sense, think, and act in an environment made up of other machines / humans to achieve a certain purpose," he then identified three areas of tunnel vision in AI research. The first: we've recently ignored building world models into AI in favor of throwing large amounts of data at a problem, letting the system develop a model itself. The second is that we currently focus on solutions that work at a point in time, rather than on solutions that continuously learn. Third is that we tend to build monolithic AI systems, whereas our world is a multi-agent ecosystem with competitive and collaborative elements.
Host Mark Anderson recalled a recent Department of Energy conference with many industry-leading AI professionals in which the consensus about the current level of AI tech was not as optimistic as the general public might imagine. He cited as the two biggest problems in AI a lack of explainability in model results - a consequence of Rao's first point - and a diminishing return on investment.
Andrew Fano pointed out that almost all business problems get seen as classification problems. Machine-learning classification is a mature and effective tool, but it's not always applicable. It's also far from the general intelligence people worry about, or "the singularity."
To illustrate the gaps in AI knowledge, he posed two questions. First, referring to a cup of coffee held by Anderson, he asked, "Is that cup garbage?" Second, "Is it better to carry water in a shoe or a sock?" Any 5-year-old could answer the second question, he said, and not because they'd observed 10,000 attempts at carrying water with both instruments, but because they have a system for common-sense reasoning. Machine learning, which is only a subset of AI, has found great success in a relatively narrow problem space, but in its current form, it is not the panacea that some claim. Fano said firms like Accenture are starting to turn their attention to more general reasoning business problems, but it's a nascent field.
Paul Maher described several machine-learning success stories in industries including healthcare, insurance, and manufacturing. He said notwithstanding the incredible advancements in superficial or tangential areas of AI, such as cloud computing, fundamental innovation on an algorithmic level is lacking. In sum, AI as an engineering problem is well-established, but AI as a science has a long way to go.
Anderson contextualized these problems on a tech level, pointing out that almost everyone is using the same "monkey wrench": neural networks. Neural networks are incredibly good at classification problems, but the algorithms have been around for a while, and much of the recent innovation has come from increases in computing power.
Throughout the panel, it was clear that everyone using machine learning employs a similar bag of tricks, and that bag is far from anything resembling human-level thought. Humans are great at thinking about a few things on a deep level, whereas computers are great at multitasking many shallow tasks. Closing that gap will require research to take a hard left turn.
J. Augusto de Oliveira, Executive Vice President and CTO, Cypress Semiconductor; and
Moderated by John Thompson,
L-R: John Thompson, J. Augusto de Oliveira, and James Barrese
J. Augusto de Oliveira: Semiconductor companies are jumping into the car industry because of autonomous driving. Automotive industry already consumes 8-9% of total semiconductor output, and that is growing at 6-8% per year.
James Barrese:If you look at what's happening, drone technology has really advanced. You can stitch together a grid of drones to lift objects.
John Thompson:Cars don't see like we see. They use a sensor array, so how are they identifying things in the world - "That's a baby carriage, that's a sign" - what's driving the seeing in the car itself?
Oliveira: We're applying heavily the same technologies that we're discussing here: advanced neural nets with deep metrics, trained exhaustively, and ... we're still a bit far from totally autonomous driving, especially in the less developed areas in the world. One of the passions of myself and my company is more, How can I start adding value at the edge? We need to have that processing done at the sensor and not in the "brain" to decrease latency. A whole new set of challenges. The automotive industry has lots of concerns about safety; now security as well.
Thompson said there's a lot of worry about cars being hacked, and asked what is being done to mitigate that.
Oliveira:"There's a lot of responsibility throughout the distributed computing environment of a car, to secure that you have the information properly fed.... We're working closely with ... all the major OEMs, and we have architected security in our product families, meaning every controller that we do that goes automotive is not only functionally safe, but also has a hardware security model built in. It has to be architected in the system; otherwise it doesn't work."
Thompson:"I think that's a conversation that Mark [Anderson] has been having for years, that you have to architect safety and security into the system or it's an afterthought, and therefore much less valuable and useful."
Thompson asked where we stand in the development of these technologies: "Are we at a place where cars and airplanes can safely be made autonomous, or are we in the early stage of the evolution of these networks?"
Barrese said he likes the metaphor of the space race: "While it feels like we're all advancing, I think 20 years from now we'll look back and go, 'They were using slide rules!'"
Oliveira:If you need to get to autonomous driving, perhaps you need to leave the roadway days, but having a car-to-environment connectivity can dramatically improve the speed, congestion, and so forth. [Autonomy] will come, but the industry isn't waiting for it.
Barrese:I agree.... I think we'll see it more in the trucking and shipping industry first, where I still need a human involved to handle things that aren't solved yet, and yet I could string together a convoy ... of 10 autonomous trucks that are all communicating, with one driver in the front. That kind of opportunity is emerging. Also, you'll see a lot more car-to-city infrastructure: red-light optimization, traffic optimization, hazards....
Thompson asked what the panelists thought would be the most important innovations in sensor technology in the next couple of years.
Barrese:I'll bring up something that I don't think people really think about. Everyone's talking about the car "seeing." We also really need good sensors inside the car. So, imagine you're a new York taxi driver and you're driving somebody home at 2am out on the bar scene, and they pass out. Do I need to take them to the hospital, or just help them wake up and get out of the car?
Oliveira:I think the most exciting thing for me would be mandating autonomous braking. That technology has the potential to save an incredible amount of lives, meaning the car automatically stops because it senses an imminent crash with the car or object in front of you. It's cheap, it's effective, and it has a great potential to save lives.
Kimberly Prather, Distinguished Professor and Distinguished Chair in Atmospheric Chemistry, Scripps Institution of Oceanography, Department of Chemistry and Biochemistry, UC San Diego;
Moderated by Larry Smarr, Founding Director, California Institute for Telecommunications and Information Technology (Calit2)
by Evan Anderson with Sally Anderson
L-R: Larry Smarr, Kimberly Prather, Rob Knight, and Jack Gilbert
You get to become some sort of microbiome gypsy, where you cross our palm with silver and we tell you your future. - Rob Knight
We hear "Don't swim," we hear "Don't surf," but we don't hear "Don't breathe." - Kimberly Prather
You are literally emitting a little microbial cloud. Think Pig-Pen from the Peanuts cartoons. - Jack Gilbert
Moderator Larry Smarr opened this session by describing microbes as largely invisible, basically "information creatures," because they're DNA. He then asked Rob Knight: "How do you go from a sample to actually 'who's there,' and what are their capabilities?"
Knight said they're creating a revolution at UCSD in their ability to understand the microbial world, both around us and in our own bodies: "Each teaspoon of your stool has just in the DNA of its microbes the information that would take a ton of DVDs to encode.... We get samples from anywhere - the human body, the ocean, the soil, even the sky - we extract the DNA from it, and we use high-end sequencing equipment to read out huge numbers of fragments of DNA from anything that was in the sample." They then piece them back together to see what microbes were present in each environment and do readouts on all the chemicals present, "so we can understand not just 'who' is there, but what they're doing."
They have the capacity to process about 100,000 samples every year, said Knight. One exciting frontier is using software to not just understand the moment, but also to predict the future. He gave examples of using a sample from a 3-week-old child to predict "how fat that kid will be at 10 years of age; we can predict how well they're going to perform on cognitive assays," as well as having fecal samples donated in 2002, along with all those donors' clinical records, so that 17 years later they can predict what their future became.
"You get to become some sort of microbiome gypsy, where you cross our palm with silver and we tell you your future," said Knight. "That would've sounded completely crazy five years ago, but the idea that from a technical perspective you can do it is looking much more plausible." Ethical questions come into play, but our microbiome can change during our lifetime: "The idea that you can take your microbiome and reshape it to your health is looking more and more plausible. Now we need a user interface for it, so it's not just a black-box idea."
Smarr then introduced Kimberly Prather as "an amazing scientist and just a force of nature," adding: "She has totally freaked me out. We all love being here, and how healthy it is to be here, next to the ocean - I thought. Kim, tell us about that?"
Prather first, good-naturedly, asked if he wanted the truth, then continued: "Okay ... Rob talks about how microbes influence human health. They're also trying to influence planetary health, and largely they're doing that through the ocean. A lot of people don't really think of the ocean; you think of the salty air you're breathing."
But their large facility allows them to isolate the ocean and study it. "They don't call it the living ocean for nothing," Prather said. "When waves crash, they put bubbles down into the ocean, and you get enrichment of bacteria, viruses, proteins, etc., in the air.... One of the things that we're showing is that microbes from as far away as Africa are changing the snowfall over California. They come all the way across."
Prather "flies through the clouds" to sample what those clouds are made of, in an effort to understand "which ones get out and influence our weather, which ones get out and make it into our bodies."
Meanwhile, millions of gallons of raw sewage are constantly going into our oceans, as well as "extreme pollution runoff events, which get worse with global warming. "We hear 'Don't swim,' we hear 'Don't surf,' but we don't hear 'Don't breathe.'"
Jack Gilbert, asked to describe his studies, said: "Every single one of you is emitting about 38 million bacterial particles into your immediate vicinity every hour that you sit down in this room ... You are literally emitting a little microbial cloud. Think Pig-Pen from the Peanuts cartoons."
Those who grow up in a farm environment breathe in things from the animals around them in addition to just human microbes. Gilbert and his team worked with both Amish communities and a Hutterite community to determine why the Amish have virtually no allergic disease: "There's no asthma, they don't really have any autoimmune disorders; and the Hutterites have these very elevated levels."
They learned that the microbes they isolated from the bedrooms of the Amish "accumulated from the farm throughout the day, stimulate their lungs into an immune reaction, which keeps their bodies from overreacting to other microbes." Changing that relationship can have a notable effect on the immune system, promoting asthma, food allergies, and skin diseases.
Gilbert then described rodent studies that corroborated their findings. Promotion of "the right" bacteria in the intestine can actually produce chemicals that can reshape many of the organs in your body and many of the cell processes. "So, yes," he said, "we can live in an environment where there are no plants, there are no animals ... but maybe we can take that farm, package it up in a drug, and reintroduce it into children that cannot get exposure to a farm." He continued with more colorful ideas that SNS members will see in a future transcript.
Gilbert contends that we now live too sterile a life. Before vaccines and antibiotics and public health works, society's main threat was infectious disease. We've since eradicated that potential, but we've gone too far. "The immune system needs that learning; it needs that training from the world around us."
Smarr: It's like your immune system is training on the world around it because that's its job - it's got to protect you from that world. If it doesn't know what the world is, how can it do that?
Knight:Exactly. It's like we didn't learn the lessons from Silent Spring, and we're creating that situation in our own households, and even inside our own bodies. I think, Kim, you were saying just before this that even in the oceans we're depleting those microbes with antibiotics.
Prather:When we even put out treated wastewater, like we do up and down the coast, we're releasing pharmaceuticals and all these sort of antibiotic drugs into the water.
Smarr:Birth control, estrogen-disrupting chemicals ...
Prather: ... Cocaine, methamphetamines, all of that. And these are all hydrophobic things. They don't like water - and so we're getting all that into the air. But it's also reshaping the natural microbes that have been evolving for over 3 billion years, and they have a method to their madness for this planet. So, yeah, it's a bum deal.
Ed. Note: Random excerpts lost in sequence but that just shouldn't be missed (see future for full transcript):
Prather: A couple times we'd been flying very carefully through these super-cooled liquid clouds looking for the magic ice crystals, and our plane quickly became the ice nucleus. And so we fell from the sky.... I'm trying to think about ways now... we can use artificial intelligence, and how can we use swarms of UAVs.
Gilbert: The first patent I ever produced was to take those bacteria that nucleate ice and add them into ice cream. So they sent me down to Antarctica to find those bugs, not into clouds - and we found bugs that make ice cream taste smoother, which was the noblest thing I've ever done.
FiRe Chair Mark Anderson (L) presents David Morris with the first-ever FiRe Herculean Award in recognition of his many years and many means of FiRe support, and for making even the most harrowing of FiRe feats seem seamless.
In Defense of Democracy: How Taiwan Stands Up to Xi Jinping
H.E. Ambassador Abraham Wen-Shang Chu, Director General, Taipei Economic and Cultural Office in Los Angeles
Hosted by Evan Anderson, CEO, INVNT/IP
In an illuminating panel, H.E. Ambassador Abraham Wen-Shang Chu and host Evan Anderson, CEO of INVNT/IP, discussed the Taiwanese struggle for democracy and freedom in the shadow of mainland China.
The ambassador explained that in 1911, Dr. Sun Yat-sen wanted to bring American and British democratic values to Taiwan. However, the government in mainland China wants as much power as possible, so it has always been a struggle. "We are mavericks," he said, "a subset experimenting with democracy." Ambassador Chu places some blame for the current situation on Confucian philosophy: "Emperors love Confucian philosophy, which hardens a non-exchange between social classes. It has the veneer of peace and harmony, but in reality, it's a form of control. Without Confucius, there would have been a cultural enlightenment 2,000 years ago in China, and it's a pity that his ideas came to prominence."
Taiwan proves that democracy can thrive in Asian countries, whereas Xi Jinping pushes the narrative that a communist government is more suitable to Chinese culture. According to Anderson, Xi has changed the face of Chinese leadership more than any figure since Mao Zedong. Anderson hopes that Xi's power will lessen as he ages, given how hard it is to maintain top-down power structures. As he put it, "Heavy lies the crown."
When asked how America has helped defend Taiwanese democracy in the past, Ambassador Chu credited their support for lack of major conflict with China over the last 70 years. Yet, he said, democracy [there] is a new thing, and Americans are lucky to have grown up with it. In contrast, Ambassador Chu said, "we've had to pay the price for democracy. There is no negotiation with communists, and you must show your strength." In response, Anderson pointed out that sometimes Americans get cynical about defending democracy abroad, but that Taiwan is a success story and a shining example of what is possible in the region.
Asked about the future, with the PRC looming overhead, Chu confidently expressed that Taiwan is the future of China. Free speech is natural, and if the mindset of the People's Republic of China prevails, then war with the US is inevitable. It won't be easy, he said, but the only way to convince them will be a show of strength; reason will fall on deaf ears. "If China is such a good country, why are there no pioneering scientists? Why are all the good ones educated in the US? There is no free will to think or invent anything. Everything is copied."
Anderson pointed out that over 1 billion people in China are unable to live up to their full human potential. That's a huge opportunity cost for society, and we could sure use their help with our larger goals, he said.
Ambassador Chu said that Americans can help the Taiwanese maintain their freedom by doing business with them: "We are honest and hardworking people. We are a little giant."
Harri Hursti, Founding Partner and Hacker, Nordic Innovation Labs; and
Moderated by Jody R. Westby, CEO, Global Cyber Risk LLC
If a foreign nation were staging a land invasion, you wouldn't ask the local sheriff to defend the fort. But this is what we do in our elections. - Harri Hursti
Moderator Jody R. Westby of Global Cyber Risk led this panel with renowned hacker Harri "scofield" Hursti of Nordic Innovation Labs and Thomas Aidan Curran of Ory, on cybersecurity vulnerabilities in the US election system.
L-R: Jody Westby, Harri Hursti, and Thomas Aidan Curran
As Westby pointed out, there are four steps with potential vulnerabilities: voter registration, machine voting, tabulation of votes, and the reporting of votes. In contrast to what most people assume, Hursti said there is no permanent IT / cybersecurity staff employed by election offices; all of that work gets outsourced: "If a foreign nation were staging a land invasion, you wouldn't ask the local sheriff to defend the fort. But this is what we do in our elections." Election teams have no security experience, and yet they're tasked with defending our votes from foreign interference.
Hursti said voter registration in the US is peculiar because voters' party affiliation, along with other personal data, is publicly available to any bad actors. He described this as "identity theft in a box." Also, voter registration databases are purged - some 60 million people were removed from the database between 2014 and 2018. Both the registration and the purges are carried out by private companies hired by local election officials.
In regards to voting machines, Hursti said that only three companies account for 85% of the market - again, hired by local election officials. While the intended purpose of spreading out election responsibility to local districts may have been safety through diversity and decentralization, it turns out that the opposite is true. Not only are there commonalities between the machines produced by these companies, but many voting districts also end up outsourcing all of the election management to one of a few private companies. These firms are high-value targets for anyone trying to hack a US election.
After voting is over, the data is often sent to county offices wirelessly over the mobile network, and other times physically delivered on a memory stick. Vote reporting used to be done internally by county officials, but increasingly, private companies are handling this step, often with little visibility. In 2007, Hursti published a 400-page report on election vulnerabilities; most of the problems he identified have yet to be fixed.
Thomas Aidan Curran firmly believes that open-source technology is the way forward for election cybersecurity. Open-source environments encourage a peer-review process that serves to improve the software and increase confidence in it. Among the missing components in the current election technology that Curran would like to see implemented are open-source voting machine software and cloud architecture, rather than hardware device architecture.
Curran pointed out that even if someone like Hursti finds a vulnerability, they wouldn't be able to patch it. Ultimately, the company producing the software has to take responsibility for the fix, and in his experience, they often don't. He added that a voting machine doesn't need to be sophisticated and proprietary, and much of the actual voting architecture would be more effective if it were in the cloud.
Curran and Hursti agreed that the primary goal of a voting system is to provide the ability to audit a count, and the best way to do this is a publicly funded and open-sourced platform, ideally with paper ballots.
Westby said that almost a billion dollars was spent on election security in two payouts to local and state election boards, boasting almost no accomplishments. She recommends that more strings be attached to any future funding.
Films Changing the World
Andrea Crosta, Executive Director, Earth League International, and subject, Sea of Shadows;
Moderated by Sharon Anderson Morris, CEO, FiReFilms
L-R: Sharon Anderson Morris, Andrea Crosta, Elizabeth Unger, Judy Korin, Lindsey Keys, and Chris Hegedus
Moderator Sharon Anderson Morris, CEO of FiReFilms, welcomed this year's esteemed panel of documentary filmmakers, then told the story of FiReFilms' start in 2012, when at the Sundance Film Festival in her home of Park City, Utah, she saw some documentaries she believed were crucial to share with the world. That year, she brought the (later) Academy Award winner The Cove and director Louie Psihoyos to FiRe (for its second-ever public screening). Every year since, there's been a FiReFilms screening night at FiRe featuring the current year's chosen documentary and its director. This year's selection, Sea of Shadows, had been screened to the FiRe audience the previous night.
Andrea Crosta, executive director of Earth League International, was introduced by Anderson Morris as "the guy who went out and risked his life many, many times, and is continuing to do so, to save a species of porpoises," the vaquita, whose numbers in the Sea of Cortez are now estimated to be down to 15 individuals. The subject of Sea of Shadows, Crosta said he'd earlier been involved with the same production company and director - Richard Ladkani - with the documentary The Ivory Game. At that time, he realized they would have special fundraising and teaching challenges, being both a smaller NGO than most, and somewhat hard to describe. ELI's solution has been to work together with important media partners.
After a clip from Sea of Shadows, Anderson Morris addressed the audience: "You guys are all world changers. Showing you the right films can change, and has changed, our world." Since the night before, she'd received three emails from people wanting to host private and corporate screenings, as well as help with an immediate need for funding to continue work on the ground. The vaquita killing season starts up again in November-December, and "we're already working connecting governments together to save this beautiful species," said Anderson Morris.
Crosta added a crucial point: "The vaquita is a flagship, is a symbol, for the whole Sea of Cortez. The moment the vaquita is gone, then others will start going there [for illegal captures], and then the whole Sea of Cortez will become a killing field of illegal gillnets to catch everything; it will be the end of a beautiful place."
Sea of Shadows was scheduled to begin airing on National Geographic channels starting November 7-8 in the US and is currently available on Amazon Prime. The project itself is a continuing effort. Since this panel, FiReFilms member Bill Soward offered to match any donation up to $20,000 to ELI for the urgent effort to save the vaquita from extinction this winter. Contact email@example.com for donation details.
Elizabeth Unger is the director of Madidi (working title), named for Madidi National Park in Bolivia's upper Amazon basin. Unger's biology background includes extensive experience in wildlife conservation. While in Bolivia in 2015 for another project, she had a tip from a government staffer that there was a huge jaguar trafficking story no one was talking about, "with a ring of Chinese nationals trafficking the parts out to China."
Unger's interest intensified when she discovered that China was investing $250 billion in Latin American infrastructure development. In other words, "Chinese companies and Chinese workers are coming into Latin America and building up projects like dams and bridges and highways, which are actually deforesting the very areas that jaguars live in. And to boot, some of these Chinese workers are opportunistically baiting local hunters to hunt jaguars at a massive rate." The media tends to be focused on elephants and rhinos in Africa, while there's no awareness that in South America "there's a decimation of a brand-new species going on, which is the jaguar."
Fast-forward four years: Unger has been working with the UN for the past two years, as well as with Leonardo DiCaprio's production company, Appian Way; and has received early support from National Geographic and Sundance. But she still needs more funding to help take the film all the way.
With Madidi in post-production, Unger has submitted to Sundance and is looking for financing to reach the end. If accepted, she'll have 10 weeks [from FiRe] until picture lock; financial support will be used to keep up momentum in the edit. She also seeks continuing support to help stop the decimation of jaguars, as well as to hold the Chinese government, Chinese companies, and local South American governments responsible for the large-scale political-economic deals being made.
Producer Judi Korn said that The Great Hack is about "a different kind of extinction - the extinction of civil discourse and democracy." She joined when Jehane Noujaim, who made Startup.com with Chris Hegedus [also on the panel], and Karim Amer, who together made The Square, were transitioning the film from initially being about the Sony hack, "where there was this incredible intersection between technology, politics, and entertainment."
As if that weren't compelling enough, then came the hack of the Democratic National Committee during the 2016 US election cycle. As they looked further into that story, they ended up shifting direction. "The Hack was not about any kind of computer system," they realized - "the Hack was really happening to people's minds - the manipulation of voters, not the manipulation of the vote itself - although that is a huge problem."
With a background in animation and graphics design, Korn's call for the film was to make the invisible world of personal data visible, in a way that wasn't dry or abstract.
The Great Hackis now on Netflix in 190 countries in 23 languages. "There's ironically a vibrant conversation happening in social media in every language," Korn said. "People are scared and worried - and worried about, more than anything else, democracy and the integrity of free and fair elections." She said The Great Hack is being called "the best horror movie of 2019."
Following a clip, Anderson Morris asked Korn what she hopes will happen as a result of people learning more about privacy issues. The filmmakers have now shifted toward doing impact work around the film and hope to continue to "inform and move people to do what they can personally to protect their privacy, to advocate for privacy, and to work to contain disinformation," said Korn. "And most important, to hold the technology platforms to account and have a real conversation about democracy, and if that's more important than this opaque advertising technology system that's been co-opted to manipulate voters."
The Great Hack is now available on Netflix. Because Netflix doesn't support impact campaigns, Korn said, they're currently looking for funds for that purpose. Korn also invited the FiRe audience: "If you'd like to have a screening, especially if you're a tech company and want to talk about the ethics of this space and personal data usage with your team, we'd love to see the tech sector come back to those idealistic roots about doing no harm, and maybe shift to 'Move slowly, Move deliberately, and Fix things.'"
Note: The Great Hack is planned as the Marquee Film for members of FiReFilms.org in November-December.
Growing up in the woods of upstate New York, Lindsay Keys, director of The Lyme Trials (working title), had Lyme disease several times as a child. She was cured each time, but then became very sick in her early 20s. She was bewildered then by the responses she encountered in doctors' offices. Lyme seemed so controversial then that "some doctors said, 'I wont even test you for that disease; don't talk to me about that.'" In one case, she was "fired" for requesting a Lyme test, even though she had all the known symptoms. Another doctor refused to test her because "I see 10 people a day like you."
"I basically unraveled," Keys said. "Physically, emotionally, mentally ... I couldn't read, I was getting lost in New York City on my way home from work, I couldn't hold cameras, and I had doctors saying 'Oh, well, you're getting older' - and I was 25! So, what is going on?"
She eventually saw a doctor who didn't accept insurance, which she thought strange, and was given a kind of test she'd never heard of before, for which she tested positive for the first time in 10 years, and thought that was strange. She was about to start a "very hard-core" experimental treatment when the nurse practitioner asked her how she was going to get through this. As Keys described it, "I said to her, 'I'm going to make a documentary about Lyme disease. I don't know how, but it's happening.' And [the nurse practitioner] said, 'Oh! We have another patient here who's your age, and he has Lyme, and he's a filmmaker. Do you want me to connect you?'" That was four years ago; since then, she and the man who became her co-director "set out on a quest that's taken us down a rabbit hole that we never knew existed."
Following a clip of the film-in-production, Anderson Morris said: "You're creating global awareness about something that people didn't really acknowledge [even] in the past - and with global warming, it's becoming more and more prevalent, correct?"
Keys answered: "Yes. And there's been little to no scientific progress in 44 years, which is another problem. I find it highly suspicious."
As the panel's final guest, Anderson Morris introduced renowned documentary filmmaker Chris Hegedus, whom she'd first met when Hegedus came to FiRe as the director / producer of Unlocking the Cage, FiRe 2016's Featured Film. Anderson Morris asked Hegedus to describe to new filmmakers what they might be able to learn from her experience of being an expert over the last couple of decades.
We thought it valuable to post Hegedus' full response (slightly edited for space):
I had a really amazing time when I was here with Unlocking the Cage, and for just what this whole community brings. For us filmmakers, it brings a lot of things to think about. And that's what filmmaking is. And I'm happy to be here with Lindsay [Keys]; I'm a producer on her film.
You know, films are very personal. I don't think there's a one-thing-fits-all. First of all, it's an art form - and it's a business. In order to do these things, I think the most obvious things that filmmakers need is money and access, and to get their foot in the door ... and be passionate! Be very passionate about what they're doing, because these films take a long time, a lot of times, and are not easy.
For the type of films that I've made over the decades - I tend to do less "issue" films than films about people who are passionate and taking a risk to make a change in their life, or in other people's lives - the stories I do are followed through people. In most of these films, I see that as well, that they're following characters who are putting themselves and their lives on the line to do something. And in the end, I think what's very special about documentaries is that they're also the history of our times. That's what we're showing people, and hopefully they can be preserved and remain an archive of that.
Then came a surprise ending -
Anderson Morris:And in case you hadn't heard, Chris was nominated for an Academy Award for The War Room. You've probably heard of Startup.com - that's her. Unlocking the Cage. And because you are such an amazing woman, we would love to take this time - it's your turn, my friend -
Hegedus: Oh no.
Anderson Morris: - to give Chris the award that she deserves so beautifully: our Biggest Global Impact Through Film award. This is for you. You deserve this.
Hegedus:That's so nice of you. Thank you so much. I took a picture the other night of Sharon with us, because she's such an advocate for filmmakers, and what she's done here with FiRe is really amazing. I'm very happy to be part of this.
Anderson Morris: Thank you; thank you. Thanks to [the whole panel] for their hard work to show us about some of the most important things that are happening in this world.
Anderson Morris closed by saying FiReFilms provides members 12-15 film links per year with a sheet on why the director made the film: "We pay the directors for the opportunity to show this to our members, which is absolutely unheard-of. People always think that directors are rich. Directors actually lose between $500,000 and $1 million per film, so FiReFilms is here to keep supporting documentary filmmakers and keep them going."
Scalable Technologies for the Climate Crisis
Aaron Fyke, Founder and Managing Partner, Thin Line Capital;
Moderated by Cynthia Figge, CEO and Co-Founder, CSRHub
by Evan Anderson with Sally Anderson
L-R: Cynthia Figge, Aaron Fyke, Stephen Honikman, and Ali Douraghy
Solutions have to be actionable by the individual. - Stephen Honikman
There needs to be more focus on agriculture, land use, and industrial sources of greenhouse gas emissions ... I don't have good news for those sectors. - Aaron Fyke
If cows were a country, they'd be Number 3: US, China, and cows.... The plant-based diet is not just for hippies and animal-lovers anymore. - Ali Douraghy
Asked by moderator Cynthia Figge to describe the state of the world, and how to characterize where we are in terms of holding temperature rise to 1.5 degrees Celsius, Aaron Fyke began by saying "It's very easy to get depressed - but I don't think people realize how much has been accomplished in the last 10 years. There's a common refrain that humans are terrible at understanding exponential growth; that's true and remains true."
As an example, he brought up the climate carbon wedge concept out of Princeton. There were 15 wedges, each 1 gigaton of carbon to be mitigated. "If we did it all, we could save the planet, [but that] seemed really impossible to achieve. The point was to give people hope, but I read all this stuff and I felt very concerned."
Then, in anticipation of the panel, he reviewed the wedges. Two refer to wind and solar. In 2007, the goal was to have "100x the amount of solar we have on the planet. That would save a gigaton of carbon by [approximately] 2050." In 2007, there were 7 gigawatts of solar, so 100x, or 700 gigawatts, was a tall order. But today we're at 500 gigawatts of solar worldwide.
"We actually are almost hitting that way ahead of schedule," Fyke said. "Wind was 10x - we needed 1,000 gigawatts of wind worldwide. We're at 600. So both of those wedges, astonishing progress." He attributes the progress to huge drops in cost for both wind and solar, both in the 70-percentile. And "LED lighting has dropped 94% in cost in the last 10 years. These are huge changes which have enabled huge deployment. So I remain hopeful."
Figge asked Stephen Honikman what "scalable" really means.
Honikman said he talks with customers one-on-one every day. "It would be great if those one-on-one interactions were scalable. They intrinsically aren't, but ... that's how it starts to spread, and the ideas of cascades and accelerating growth."
He'd like to blend the concepts of "thinking globally and acting locally, for any application." The essential step of reducing cost brings up other issues: "We're going to make sure the scalability is sustainable, too, but to have something spread in a sense that we'd all agree is 'scalable,' it has to have a global impact, and it has to be achievable by the individual." "The technology is there," he said, "this is now actually a human issue. To get climate solutions to scale, whether it's young people or citizens in mainland China, you must cater the discussion. Solutions have to be actionable by the individual."
Figge then turned to Ali Douraghy, quoting him as saying it's not just what we're working on, it's also what we value.
Douraghy's focus is on things that won't go to scale until closer to 20 years out. "The wind and solar piece definitely has been part of the equation, and then.... It looks very different than it did 10 years ago, because you have an entire generation that has grown up in the context of climate change. And the way that they view 2050, which is a critical year in our heading towards zero emissions - they view that as something that's very much part of their life span."
Figge said the language has shifted to really using the term "climate crisis." Some in the corporate sector are setting "some pretty audacious goals, but they tend to be at a time horizon between 2030 and 2050." She asked if that's legitimate anymore. Addressing Fyke, she asked: "What are a few scalable technologies that we're not talking about as much, but they're here and fairly ready to go?"
Part of the challenge is that the problem is "dominated by large industrial players - Ford, Dupont, GE and GM - not startups," said Fyke. "And part of the reason costs have fallen has nothing to do with 110 venture-backed startups; it was the big manufacturing machine of China turned on, and solar costs collapsed. Still, many people think of energy generation (wind, solar, electric cars, etc.) as the problem, when that's only maybe half.... There needs to be more focus on agriculture, land use, and industrial sources of greenhouse gas emissions ... I don't have good news for those sectors."
Asked what's going on in his industry, the microgrid, Honickman said challenges with the grid are fairly clear: "The deal we made with the devil between government and big capital 100-odd years ago was, 'We'll give you the keys to the monopoly to sell power, and no one can compete with you, but we're going to regulate you. And the one thing you have to do - utilities - is deliver power, and reliably.'" Both of those are questionable today, while technology has kept advancing - so the need back then for "a big central power station and just direct lines out to the end users ... fundamentally doesn't make sense today."
Energy generation on the grid is about one-third of industrial emissions. "A trillion dollars have been poured into this incredible machine called the grid.... and the grid would be better off if it wasn't a single point of failure." He added: "Individuals have the ability to take agency over their energy use and budget." Microgrids would be a big step in adjusting the problem the grid represents.
Figge asked whether politics at the federal level can keep us from moving forward, "and if that's true, can efforts at the state or city level achieve scale?" Douraghy said he's very encouraged that the reaction by many state and city leaders to the US pulling out of the Paris Accords has been signing the pledge with the tagline "We're still in."
Federal leadership is causing concern on an international scale, and absent of US leadership, there's no good second alternative. "When we're in the game, we can bring allies from all over the world on board," said Douraghy. "Absent the US, there is not the same coordination. We need to be in that setting as well."
Douraghy added that the individual also has powerful leverage via their food choices. In terms of land use and methane emission, "if cows were a country, they'd be Number 3: US, China, and cows.... So it is a big deal. The plant-based diet is not just for hippies and animal-lovers anymore."
Anne Hardy, Chief Security Officer, Join Digital Inc.; and
Moderated by Greg Ness, CMO, Tempered Networks
Edited excerpts by Evan Anderson with Sally Anderson
You can't protect something you don't know is there. - Steve Fey
In our world, there is no border, there is no wall; we are all connected all the time. - Anne Hardy
Greg Ness: About 10 years ago, all sorts of companies started connecting devices to the internet. Some of those devices are running firmware that has never been updated or developed, so every device you have in your life (your laptop, your phone) is likely 10 times more secure than some of these devices that never had updates and were never designed with any kind of security in mind. You now have billions of devices creating a massive attack surface, and then access between the devices creating an even bigger problem, which is the attack vector sprawl problem.
L-R: Greg Ness, Anne Hardy, and Steve Fey
We've had devastating attacks in the past few years. One was NotPetya, with Russians attacking Ukrainian infrastructure. Oops - it spread all over. It affected Maersk Shipping, Federal Express, took down hundreds of hospitals, all by accident because of the attack surface and the vectors.
Let's talk about IoT and IIoT. How are they different?
Steve Fey:The internet of things and the industrial internet of things. There's also OT, operational technology - the broad umbrella over IoT. We're talking about the control systems in a factory or a building. They went digital in the '80s; in the '90s, they started being networked. Today they're all networked, some of them wireless, and they are accessible over the internet.
Ness:What are the implications of this, and how does our security thinking need to change to address this issue?
Fey: In enterprise, cybersecurity has been an issue for decades, on some level. Spending is about $150 billion a year. When it comes to IoT, it's as much a cultural challenge as it is a technical one. The companies that build IoT systems are thinking function first. No one thought they'd be hacked, but now it's routine: the latest statistic I've heard is that they put an IoT system online, and within 15 minutes it's being discovered. When ransomware hits a building, it all stops. They can keep it up, but to rebuild that server can take weeks, and about $75,000.
Ness:What has really raised awareness around this issue?
Anne Hardy:Well, first, it's incidents. And then the fact that we have pretty much billions, now, of devices spinning out - in buildings and homes - and I think it disrupts the way we've been thinking. It seems that the hackers have found a way to get around the infrastructure to get to these devices, while we, as IT people and technology people, have not really changed our thinking.
Ness: People often ask, How did this happen - and why, with all the money spent, isn't it better? Most of the firewalls in IPS systems, etc., were architected to address '90s-era scenarios and then TCP/IP. We have culture issues now.... Instead of one single drawbridge around that castle with the moat, you now have almost an infinite amount of drawbridges. The network security vendors love the idea of selling you a firewall for every five to ten attack vectors. And they've almost evolved into a pharmaceutical: "We want to treat the symptom and monetize it."
Hardy: In terms of the culture, it's not just about educating people. We have to come up with solutions that have security by design. It's no longer like a band-aid, where you fix it after the fact. In our world, there is no border, there is no wall; we are all connected all the time. Everything is invisible, so you have to really think about security. Whatever you do, security is never going to be perfect. We have to be ready for bad things to happen, because something bad is going to happen.
Ness: One hundred seventy new devices are being connected per second. In terms of device visibility, how much of this do we know, versus how much we only kind of know?
Fey: We're in commercial real-estate security. When you ask the largest building owners, "Do you know what systems you have in your buildings?" the answer is "No." They really don't.
You can't protect something you don't know is there. You also have to know who can talk to these systems. Who has access? If you own 100 buildings, chances are you have about four or five network control systems in those buildings. That's over 500 systems you need to know about. Among those, you probably have over 200 different vendors. Who keeps track of what those vendors and their employees are doing?
Ness: Now we're seeing variants of NotPetya that aren't designed for ransomware, but for pure destruction. There's no sanctity for human life: if you attack a power station, you literally could kill people. What do you see as the most disturbing risk scenario evolving from this situation?
Fey: Life safety is the biggest concern. If you can shut down an entire data center by getting into the environmental system, raise the temperature in a server center and all the servers start failing on you, that's a pretty catastrophic situation.
Hardy: As we become more dependent on all these devices, it's pretty probable that hackers could just shut down countries. We've already seen it done with cities. Entire systems that cities rely upon are just gone from connectivity; that could be the same [for countries].
David Brin, Physicist and Sci-Fi Author: Do you think things would be better if we simply reduced, by an order of magnitude, the number of bots that are using American computers and serving as base of operations for many of these plagues? In other words, if the next president of the United States were to say, "It is a matter of national patriotism that all of us simply get the latest bot removal software" -
Ness: I'm not sure, in terms of industrial devices, how much of that could be accomplished, especially with the stuff that can't be updated.
Hardy: I wouldn't say it's enough. Bots are a good way for people to get stuff done, but there are other ways we've seen at attack points.
Ness: Distributed denial attacks from bots, but they could come in through these vectors without having to even do that. They could use a single compromised device.
Fey: In the studies that we've looked at, anywhere from 60 to 90 percent of all breaches are the result of We the People doing silly things, through ignorance, not even being aware of what we should be doing. This is us. It's getting us to take this seriously and having good practices. Change your passwords on a regular basis - those are the kinds of things that are going to make a difference.
Mary Branscombe, Independent Tech Journalist, calls out: No, don't! Passwords are like shoes - change them when they wear out! [Laughter]
Longtime FiRe family member and friend Russ Daggatt spoke from the heart in a tribute to the recently belated Sidney Rittenberg. Russ introduced us to Sidney and his wife and partner, Yulin, in the earliest days of FiRe; in fact, Sidney spoke at the very first FiRe, in 2003. If you saw him onstage, in the corridors, or on the lawn, then or in any of many years after that, you will remember his flashing wit, deep heart, unparalleled memory, and double-bladed sense of humor.
Russ began by saying, in typical Russ-humble fashion: "My one significant contribution to FiRe over all the years I've been coming was introducing Sidney to FiRe and to the Anderson family, who came to love him just as much as I did." You can learn more about Sidney at the link above, or as Russ said: "Just Google him - his obituary was on the front page of the New York Times." That alone speaks volumes.
Sidney passed in August at the age of 98. He will always have a seat at our table.
Pamela Brody-Heine, Director, Clean Electronic Production Network, Green America Center for Sustainability Solutions;
Introductions by Berit Anderson, Director of Programs, SNS and FiRe
Edited excerpts by Evan Anderson
Six-minute views of the future
Pamela Brody-Heine, Director, Clean Electronic Production Network, Green America Center for Sustainability Solutions
We've talked about devices being everywhere. I want to talk about the tens of millions of workers who make those devices, and how to improve their world.
Three Things: Clean Electronics Production Network
1. Who we are: We're a program of Green America Center for Sustainability Solutions. We bring together groups of stakeholders to solve supply-chain sustainability issues that can't be solved by one group alone. We work with industry, NGOs, worker representatives, academics, etc.
2. What we do:Our goal is to move to zero exposure to toxic chemicals for workers in the electronics sector. There are two things we can do to protect people from exposure. One is to eliminate the pathways for exposure. Today, though, we're talking about how to empower the workers and inform them about what they're exposed to.
3. Why it's important: Chemicals make up the backbone of the materials in the electronic products we use every day. There are tens of millions of workers who are exposed to toxic chemicals as they make our devices. According to the UN special rapporteur, it's estimated that one worker dies every 15 seconds from toxic exposures at work. There are myriad other health, reproductive, and other kinds of impacts.
Monitoring is done, but it is far from sufficient, and workers are not informed of the results. There are new groundbreaking technologies, passive collectors, wristbands - but they take too long for testing and cannot lead to immediate, actionable results for workers.
What do we need to do? What if there were a simple way for workers to monitor their personal exposure? We need a device that's inexpensive, readily accessible, lightweight, convenient to wear, able to detect a wide range of chemicals, sensitive enough to detect those chemicals at concentrations that are meaningful for human health, providing results quickly and easily, so - maybe connected to an app on a smartphone or downloadable to a computer, able to identify the timing and location of the exposure ... and ultimately connected to a centralized system so that that information could be tracked on a larger scale.
Pamela challenged the FiRe audience to help achieve this vision to make a drastically better future for workers in the electronics industry and beyond.
Charlie Corredor, Vice President Operations, OtoNexus Medical Technologies
In healthcare, the future is bright.
First, the proliferation of sensors in small-scale spaces. Many of us have heard the word nanotechnology, which is the ability to modify atoms or small particles. Imagine a world where you have tiny particles, as thick as a hundred-thousandth of this sheet of paper. Imagine you can manipulate them on command to deliver a drug, remove tissue (cancer), and you can put them into your body through your skin.
I also think of sensitivity: how we can quantify and understand DNA and RNA. Imagine a sensor that's in your toilet, and every time you use the restroom it makes a measurement. Down the road, you can say, "Hey Google, Hey Alexa, how is my health today?" The sensors that we have now, in five years' time, will be able to do this. All of this information we gather will help us to find patterns about a truly healthy state.
Third is diagnostics. We spend a lot of time in medicine making things small. OtoNexus designed an ultrasound that fits in your ear. That's today, so imagine what is going to happen in five years' time.
Imagine a world where Alexa tells you to take a pill because your iron concentration is low, while this pill is being manufactured in a 3D printer. Guys, I'm telling you, the future is bright. Keep pushing the envelope, and we'll see you in 10 years.
Junaid Islam, Cyber Security Expert, OODA
I've been requested, sometimes by the US government, to investigate cyber attacks.
When Iran attacked Saudi Arabia in 2017, they sent me, alone, to work for the prince. We learned that Iran had developed a new type of self-propagating malware. So, you know, when people say "Don't click on that link" - well, no one in Saudi Arabia clicked on any link. All they did was turn on their laptops, and the virus was able to autonomously go from laptop to laptop. In about two days, 60,000 laptops got wiped out. A few months later, Russia launched the same attack on Ukraine, and 200,000 laptops got wiped out. So what's the correlation? Ukraine is a more open society than Saudi Arabia. The more open you are, the more lethal the cyber-attack becomes.
America is 100 times more open than Ukraine. So if the Russians launched that attack, it would be far more devastating here.
It's a very abstract idea here. What people are worried about here is what's going on with Facebook and their data. Whether or not people care about cyber attacks, if you talk to anybody about Facebook, everybody is universally upset, wherever they are on the political spectrum.
While cyber war, the geopolitical, is not that interesting to people, their own data is very interesting to [other] people. So what I predict is we're going to see a shift in how the internet is used, and it's going to be much more locked down and secure. We're already seeing a legislation shift in that direction with the CCPA [California Consumer Privacy Act] in California, which goes into law January 1st. There is now a fine of $7,500 for every identity that gets breached. If you know how legislation in America works, California goes first and then everybody piles on....
We have a juxtaposition between everybody being upset at their data being stolen and legislation coming in. I think that encryption will become widespread, and that the new internet will be private.... We are going to go to an internet where it's going to verify who you are, whether you're doing a banking app or Facebook, and it's going to lock down your data.
Quantum computing is also on its way. These new computers allow you to create new kinds of math, multidimensional computing. It's going to be even easier for governments to look at data on a massive scale, which will lead to a counter-increase in privacy and encryption. This is going to play out over the next few years. The battles you see on privacy will be the battle you see for the internet for the coming decade.
Edy Liongosari, Chief Research Scientist and Managing Director, Accenture
In a number of our sessions, the question of ethics has come up at the end again and again. I like to quote Spiderman, Uncle Ben: "With great power comes great responsibility." When you talk about powerful tools, that's what it's all about, whether it's CRISPR, DNA editing, or AI.
There are all these huge publications that offer really long philosophical dives into the ethics of AI. Who is even reading them? They're the size of the Bible. The reality is that AI is already here, people are using it - specifically machine learning. Are they going to wait? No.
Accenture clients are coming to us and saying, "Help us decipher this." Many of them are just banks, insurance companies, manufacturing, retail. They want to use AI and be more effective, but in the process they want to address the ethical issues.
What do we mean by "responsible AI"? Three things: One is transparency - transparency at a broad scale, from end-to-end of the data. Two is fairness. Data is inherently biased, so let's not talk about where the bias is - that kind of thing. In fact, the reason why machine learning works is that the data has some biases in it. But we want to make sure there are no unintended biases. The other one is accountability: someone has to be accountable for what the AI is doing. The other part is around agency. AI is about amplifying the efficiency and effectiveness of the human.
There are four aspects to AI and Ethics:
1. Organization and Governance
2. Culture and People
David Gruber, Presidential Professor of Biology, City University of New York / National Geographic Society
I'm a biological oceanographer, so I'm the one who brings all the bad news. I want to bring an upbeat note today. Watching the coral reefs decline year after year has helped me think about how, as a scientist, I might have the capacity to make some change.
We have the technology, really, to change. We probably have the technology, if we wanted to, to bring the CO2 to 180 parts per million, which is a glacial period, and go back into an ice age. But if we did do that, would we be okay? Is the solution for homo sapiens solved? No, right? They're much deeper than that. This is why I almost wanted to go get a psychology degree. This is irrational behavior; this is human psychology.
I'll just show you quick run-throughs of things that I'm doing in my science that I call "subversive empathy building," because I try to come up with projects that actually use technology to connect us to life. The hypothesis is that if we connect to life, we will see that we're a part of it, and just one of millions of species, and we'll then be able to kind of drive "Spaceship Earth" better. I don't know if that's going to happen or not, but I'm planning for it.
So, animal-eye cameras. I discovered a fluorescent shark a few years ago, and I started asking, What does it mean to be a fluorescent shark? This is more of an interdisciplinary approach of, like, looking at the pigments in the eye and using something called micro-spectrophotometry to look in there and see exactly what the visual reality of the world of that shark is.
It [the shark] is a monochromat. So even when we think about our eyes, us being mammalian primates, we design all of our technology in RGB so it suits our crazy little bodies, ourselves. So this shark, if it were to make its own iPhone, it would be a very different iPhone. Beyond that, I could then use the technology that we have to make something like a shark-eye camera, and to connect to the world, to see the world from the eye of a shark.
[Referring to slides] ... You can look up some of these papers in detail ...
Here is the shark itself. What we can see is it gives us this impression that these are the secret patterns that these sharks can see other sharks. So by getting behind the eye, by using technology, we can start seeing things from their perspective, which makes all these things start to make sense.
I also found a fluorescent turtle a few years ago that worked its way all the way onto Saturday Night Live, where they made a parody of it. Now I'm building using hyperspectral imaging, making turtle-eye cameras, because the turtle eye is so much more complicated than a monochromat. It actually has multiple rods, multiple cones, double cones, as well as colored oil droplets, which are really interesting.
So I can't just use a regular camera; I've got to use special hyperspectral and then write the algorithms to see from that world. Which is fun, because then I get to go swim with the turtles, with eye charts [audience laughter] and make eye graphs to see how this shark is seeing the world, and all these - I'm glad you liked that. Yeah, this was a fun exercise in using a mathematical modeling, is to see how a shark versus a turtle sees.
I've also been collaborating really deeply with Rob Woods' lab at the Harvard Microrobotics Laboratory at the Wyss Institute, and we've been into this idea of "gentle robots." This just came out in Science Robotics a few weeks ago, and we believe it's one of the most gentle robots ever created. It could touch a jellyfish at one-tenth of the pressure that the eyelid rests on the eyeball. So it's the idea of using technology to create gentleness, to create empathy, to connect us.
This ... Here's the ultra-gentle manipulation. We're currently doing a study - and we've got really cool results - that it does turn out that the jellyfish produces less stress. I did an entire transcript where I looked at every gene being expressed, and I saw that all the stress responses, when I would even hold the jellyfish tight versus using this gentle robot - even something as primitive, 500 million years old, as a jellyfish, does experience stress - made me totally revisit how I even see human stress.
Here's this cool jellyfish-grab ... This is another thing that we invented, called a rotary-actuated dodecahedron, which is ... the idea of going to the deep sea, but to not extract, not have this extractive mentality. We hear a lot about deep-sea mining. There are animals in the deep sea that could be 18,000 years old. Time-space takes on a whole different kind of continuum down there, and as I approach a new animal that I meet in the deep sea, I want to approach it with kindness, and I want to approach it with care.
And this ... We were able to encapsulate this jellyfish. Once it's inside here, we could use DNA swab technology, we could use cameras in here, and we could take a 3D scan and we could print that to the shape and texture and let it go....
Another one: I was inside a submarine at night, and I realized I couldn't touch anything, and I couldn't do any science in the ocean. So this was a robotic arm, a soft robotic arm, instead of a robotic claw, that goes on the front of a submarine, that could delicately touch life. There it is, grabbing. It's actually doing whatever my hand is doing.
And the upcoming one: This is just a little maybe preview for next year? But this was thinking about deep-machine learning. This is the first paper that's applying deep-machine learning for sperm whales, bioacoustics - the animal with the largest mammalian brain, totally evolutionary obscurity. It's almost - when you look at it on the phylogenetic chart, it's way on the outside. [Closes with whale song]
Berit Anderson (L) and Pamela Brody-Heine
Bill Ribaudo, Senior Partner, Deloitte & Touche LLP
Hosted by John Wells, Producer, Cool Science Radio, KPCW (NPR affiliate)
Host John Wells interviewed Deloitte's Bill Ribaudo about corporate business models and their effect on a nation's economy. Ribaudo explained some of the research his team has been focused on for the past few years. They started by exploring why some companies are worth dramatically more than others, despite often lacking cash flow.
In categorizing corporate models, Ribaudo looks at the type of assets a company uses to make its money. Asset builders manufacture physical assets with physical inputs. Service providers like Deloitte use human or financial inputs to produce services. Once you cross the "digital divide," you have companies that use soft assets, like IP. Finally, there are network orchestrators like Facebook. As you move through this list, you have an increasing ability to scale business.
In the first category, you have to build something before you sell it. Above the digital divide, there are small, incremental costs to sell more units of a product once it's created. Putting this all together, "a company is worth a multiple of its revenue, based on the assets it uses to run its business," Ribaudo said.
Since investors are looking for a promise of future revenue, companies that can exponentially scale their operation - those above the digital divide - are particularly attractive. Ribaudo then prompted: "If a dollar of revenue is not equal to a dollar of value, what else do we have to question?"
Most of the companies listed on the S&P 500 are below the digital divide. While these firms might use technology, it's not their product.
In light of his early findings, Ribaudo questioned the validity of GDP as a complete measure of a country's economy. GDP - the sum of a nation's output of goods and services - relies on the assumption that $1 of revenue equals $1 of value. If this assumption is false, how can GDP be used as a comparison between countries?
Ribaudo has found, through three separate studies over seven years, that if a country, through its regulatory policies, supports all four business models, the economy grows. Conversely, if a country focuses only on businesses below the digital divide, its economy shrinks. Ribaudo believes that a healthy balance of each category is essential for economic growth.
Ribaudo's team also researched business education to see whether schools have adapted to teaching about digital corporate models. They found that both in the US and abroad, education, even at the highest levels, focuses primarily on businesses from the bottom two categories. While Ribaudo has a strong feeling that businesses, governments, and schools need to work together to encourage diverse corporate models, especially above the digital divide, he remarked that it's currently a chicken-and-egg type of problem: traditional businesses should strive to move up by buying or building business above the digital divide, and education needs to include content on these new business models. The government needs to create an environment for all of that to flourish.
Ribaudo closed by saying: "Nations that build the greatest wealth will be those that can successfully integrate the adoption of all four business models into their regulatory policies, modernize their educational content, and provide this regulatory environment."
With host and Calit2 Lab Founding Director Larry Smarr
See all of Kris Krg's Tour photos in the FiRe 2019 Thursday PM Gallery, pp. 5-6
Hillel Cooperman, Senior Vice President, User Experience Design, Oracle
Hosted by John Wells, Producer, Cool Science Radio, KPCW (NPR Affiliate)
In Friday morning's opening discussion, Hillel Cooperman, SVP of Oracle's User Experience Design team, talked with host John Wells about the recent paradigm shift in enterprise software. When he joined Oracle, Cooperman was surprised to learn how much software they made that directly interacted with users, as opposed to just providing an infrastructure piece. Most users have no choice as to what platform they use - someone high up at their company licenses Oracle software, for example, and everyone else is subject to that decision. Cooperman enjoys getting to advocate for those people through his design decisions, making products people want to use instead of products they have to use.
Cooperman considers design as intentional and thoughtful in solving problems for human beings, and sees two challenges in designing software: the functional and the emotional. No one likes creating expense reports, for example. If polled, people might complain about how many fields need to be filled out, or how many buttons there are. But the real problem with expense reports is that you've spent money on work, and you want it back. Instead of trying to design a sleek UI to try to make this more fun, Cooperman said the solution is to scrap it entirely and make it as automatic as possible, so nobody has to spend time on it. Good design decisions are made by considering what problems users are facing, and why they need a software solution in the first place.
Cooperman also thinks software should adapt to us automatically. He gave some examples of how the same software could present varying experiences to people with different roles in the company. While he enjoys hearing that the software he designs is beautiful or easy to use, what really excites him is when his products make people feel understood.
In the early 1930s, Henry Ford famously said of his Number One car company: "You can have a model T in any color, as long as it's black." Then scrappy competitor GM came along, bringing design and aesthetics to cars. Every few years, the design team would add a new look to virtually the same car, and GM soon eclipsed Ford.
A significant disruption initially defines many industries, like moving from horse-drawn carriages to automobiles. However, as the industry matures, companies use design to differentiate themselves. Software is now at the point the car industry was in the '30s. It used to be that many good designers preferred to work on consumer software rather than enterprise, but now leaders in the field are working on incredibly complex products that necessitate great design.
While some steadfastly claim that they care only about function, not aesthetics, Cooperman doesn't believe them. He thinks the harder-to-define metrics, like how a product makes users feel, is incredibly important. Just because a product is used at work doesn't mean it shouldn't be emotionally resonant. Some companies, like Disney and IKEA, understand that design is a vehicle for telling a story, and they manage every interaction you have with their product or team or marketing campaign. Cooperman has been blown away by Oracle's receptiveness to being design-centered, rather than just redesigning its software.
Paul Jacobs, Chair and CEO, XCOM; Former Chair and CEO, Qualcomm
Hosted by Mark Anderson, Chair, Future in Review
Mark Anderson: We talked before, when things were moving toward 3 and then 4G. It was essentially a question of simple math, putting out more nodes and moving them closer together. It seems like that's what we're doing today.
Paul Jacobs:Now we're doing it slightly differently. Now we're putting out the radio, and using the fact that there's a lot of fiber and high-speed connectivity to bring it back to a central location. People call that Cloud RAN, CRAN. Now what you're going to see is lots of antennas around, but smaller things. We actually want the phone to be part of that network. So we're going to blur that distinction between the network and the phone.
Anderson: And that's where all the money math changes.
We're building a world where the network essentially becomes a community of devices. Those kinds of networks are much more resilient.
A lot of this will be built out in the developing world first.
5G is basically the same technology as 4G, but a wider band; they talk about latency and higher reliability, but that stuff isn't actually implemented yet.
The goal is that we will have massive distribution of antennas, including the phone.
This is going to help with things like sports betting in real time; when you're there, you're closer to the source.
Anderson: Mobile-enabled gambling in China would be the best money-making opportunity to come out of this.
The mobile aspect is really important in the developing markets, because there the mobile operators don't have the resources to build out the infrastructure. You'll need to turn some people into "Uber drivers" who are willing to offer their phone as a node. They get credit and their internet access is restored.
Anderson: Is there a role for satellite constellations in any future generations of telecommunications?
Jacobs: Yeah. If you try and replace the cellphone, that's a bad thing, but if you augment by putting in more geographical coverage in rural areas or places where people pass through...
Building a New Internet to Protect People and Data
The Challenge Team (L-R in photo below): Judges: Harri "scofield" Hursti, Founding Partner and Hacker, Nordic Innovation Labs; Ty Carlson, Founder and CTO, Pattern Computer Inc.; and Mark Anderson, FiRe Chair and CEO, Pattern Computer. Team Members: Richard Marshall, Chair, Cinturion Group; Steve Coy, CEO and CTO, TimeLike Systems; Robert Armstrong, Senior Software Engineer, Tempered Networks; Mark Mahan, President, MMCO; Junaid Islam, Cybersecurity Expert, OODA; Marcus Sachs, Chief Security Officer, Pattern Computer; and others not shown.
The 2019 FiRe CTO Design Challenge was "to build a new internet that will protect people and data, and will provide privacy and security for internet users, people, organizations, and entities. Provide personal anonymity when appropriate, likewise provide full attribution of users, machines, and code when appropriate. Deter the use of the internet for criminal, terrorist, or offensive military actions, optimize the internet for social good, economic prosperity, and political / religious diversity, and preserve the open concept of the internet through a suitable governance model."
The Challenge team felt they'd met all of these goals, but that their solution might differ from the judges' expectations. "The soul of the internet is a trust model from the 1970s, which is laid out in the ARPANET reference model. Trust is in the users, not the network." The team felt that changing this mindset would be too difficult, so they decided to embrace the spirit of "the trust is in the user" and aimed to formalize it with their solution. Apps are a comfortable concept for users, so they explored an app-centric internet where users can determine how they want to interact with the web by choosing which apps they use. Their additional goals were ease of use - they "want Grandma to be able to use it" - and speed of implementation.
The culmination of all these ideas turned out to be akin to Tinder for the whole internet. Just as users of the dating app Tinder can swipe to decide if they want to meet potential dates, so could users of this new internet swipe to interact, or not, with other entities. In everyday life, people can choose who they interact with, so why are they flooded with unwanted things on the internet? With the Tinder-like model, users can make verifiable claims, such as having a college degree or owning a mortgage. In this way, they can build trust when they want to interact with businesses or other entities on the web.
Airports are an existing example of such a system. Anyone can freely walk into an airport, but if they want to get through security, they have to produce proof that they belong there. They can go further by providing more information to the TSA, which gives them a better experience with TSA PreCheck. So we're already doing this, but the idea is to use it everywhere online.
After implementing this trust model, it would combine with existing software infrastructure like Software Development Kits (SDKs), web browser plugins, and Virtual Private Networks (VPNs) to change a user's experience based on the level of trust at which they're operating. For example, access to banking would be restricted. A user on a mobile device could browse certain websites without identifying themselves. If they wanted to access a banking app, however, they would have to engage a "trust me" mode, which would identify them and grant access to more-sensitive services. It would serve as a personal VPN for users, which is not a new idea. What is new is allowing individuals to dictate how a large carrier operates. Users would be able to create an encrypted connection with any other individual or machine that they want, and nothing else.
Through their research, the team discovered that there have been previous attempts by other groups, including from a US government and industry collaboration, but that it was not yet a widespread idea. The team challenged the audience to pick up where they left off and help design the low-level components such as protocols that would enable this vision.
Judge Harri Hursti [who happens to be renowned for his hacking talents - Ed.] was concerned that this model starts to put liability onto internet carriers, which has never happened before. He pointed out that many of the team's ideas relate to the Internet Protocol version 4. We are currently transitioning to version 6, which will address many of the problems discussed in this challenge. He also worried that the team conflated authentication with trust. He laughed: "Dr. Evil is still evil, even if you know it is him." Lastly, if you turn verification into a mechanical and technical process, how will you account for the increasing role that human vulnerability will play?
The team acknowledged that the presentation conflated authentication and trust, but with their platform, you could "swipe left on a well-identified Dr. Evil." Regarding the IPv6 issue, if and when that's implemented, the team believes their platform will work well with it. One of the current problems with IPv6 is a lack of economic incentives for carriers to use it. They claimed they are the first to create a system that benefits everyone. Consumers win by only having authenticated traffic that they select. Service providers win by selling products without the risk of cyberattacks. Carriers win by providing users the chance to use better routes to their service, like an IPv6 VPN, for a fee.
Previous attempts to create a system like this failed because they never ensured that everyone on the supply chain benefited from the implementation. Regarding the human factor, the team surmised that an ecosystem of third-party verification platforms to do the legwork for consumers would pop up, benefiting users unable or unwilling to review vendors themselves.
In his assessment, judge Ty Carlson said he'd hoped for a bit more of a "moonshot" idea, and was disappointed in the team's incremental change. When he worked at Microsoft, he found that whenever you make a user click a lot of buttons or read dialog boxes, they always click "ok" rather than properly engaging. He believes that security needs to be a one-time installation and start at the device level by an appropriate referee, and that a software-level solution is always going to be prone to things like virtualization and spoofing: "Security, if it's done right, needs to be invisible." Given how much our economy is going to rely on the internet in the coming years, we need much bigger ideas to be able to secure people and devices properly and have it be robust to the future.
In response, the team strongly disagreed with the notion that a single government body should be responsible for all security features, pointing out that China uses such a model. Even if we trusted our current regime, shifts in power in the government could put this technology into dangerous hands. "This is the ultimate instrument of oppression and despotism."
The last judge, Mark Anderson, expressed concerns about relying on VPN technologies, which he said have been hacked. He liked the verifiable-news component of the plan, but thought a missing use case was users who have established their trustworthiness but don't want to be identified to the site they're visiting.
Hursti gave the team a final score of 2/3, Anderson a 2/3, and Carlson 1/3, for a total of 5/9. As Anderson wryly remarked: "Not bad for an impossible task" - which is what the FiRe team believes makes the annual CTO Challenge one of the most valuable impactful in the world.
[Ed. Note: Stay tuned for the evolution of the Challenge. To see the complete wording of this year's CTO Challenge and the precursor wording re: why this solution is needed, see this paper by Marc Sachs.]
Michael Norman, Director, San Diego Supercomputer Center, and Distinguished Professor of Physics, UC San Diego; and
Moderated by Larry Smarr, Founding Director, California Institute for Telecommunications and Information Technology (Calit2)
L-R: Larry Smarr, Michael Norman, and Osh. Agabi
I'm going to make a strong statement for the first time, here at this conference ... something that has dawned on me in the last month or two. - Michael Norman
Opening on a note of computer history, Michael Norman described the start of his career in the world of Seymour Cray-engineered supercomputers with custom software and hardware - an era that gave way in the mid '90s to "what we call 'the attack of the killer micros,'" when microprocessors, particularly from Intel, started growing in precision and power.
By the end of the '90s, supercomputers were being built out of commodity processors: "COTS (commercial, off-the-shelf) clusters led us to petascale computers today, and now we're on the road to exascale. We're already at one-fifth of an exaflop, 26,000 NVIDIA Volta GPUs. Each Volta has 3,000 or 4,000 processors. The way that we've gotten to one-fifth of an exascale is massively parallel compute processors and massively parallel systems, and the future is going to look exactly like that," said Norman.
Moderator Larry Smarr emphasized that "the exascale is 1 billion times a Cray 2. Even for somebody who's worked on these systems all the way along, it's really kind of hard to get your mind around that."
What Smarr sees as being the most disruptive thing right now is that "those are single supercomputers - what if we blow up this idea of a highly parallel computer and spread it all around the world? That's what Ty Carlson mentioned earlier: the Pacific Research Platform. We now have scattered research universities around the world that can do distributed storage, petabytes, hundreds of GPUs, thousands of cores," he said. "There's the Earth Science Grid, which takes all of these things and puts them together, and it's 'potluck supercomputer,' basically ... the actual cycles per month that open-science grid provides to researchers is equal to, essentially, the cycles from all the NSF supercomputers put together."
Smarr: Mike, you were telling me some amazing thing this morning about the level of integration that's going on in computing. Do you want to tell us about that for a bit?
Norman: I'm going to make a strong statement for the first time, here at this conference, which is something that has dawned on me in the last month or two: The end of Moore's law does not matter.
Norman talked about a new startup that's marketing a processor the size of a dinner plate, the whole wafer working in concerted fashion at native clock rates, with a trillion gates. "They're using it for the fastest deep-learning processor on the market. It can learn a hundred times faster than standard approaches.... When I say, 'The end of Moore's law doesn't matter,' it's because AI is going to become an algorithmic accelerator for simulations, and it's almost unbounded in its potential."
Smarr said that the 1 trillion sensors makes him think of the human brain, "which is a biological computer." He then asked panelist Osh. Agabi: "Osh, we all have wetwear sitting on our heads, but you've been looking at a lot of different ways that we can begin to think about architecting biological computing."
Agabi said that in 2015, one of the fundamental calculations they made was that if they "took a cubic millimeter of brain and tried to calculate what that looks like on the electronic scale, when you consider the synapses in this cubic millimeter ... and try to lay it out on a wafer, you would have to build a wafer about the size of a basketball court."
"Biology has had 600 million years to essentially build a computing platform," Agabi said, "and this computing platform is super dense." He agreed with Norman that Moore's law doesn't matter, "because biology doesn't even take Moore's law seriously. It builds things on an atomically precise scale."
The trickiest part is to build a concrete business case.
Agabi said we can take biological neurons and genetically modify them to express what are called GPCRs - essentially, proteins merged with silicon.
"We have these cells - they are sitting on silicon. It has a semi-permeable membrane on it, a pump. So it's essentially collecting the air from the outside world. The only thing that we program into these cells is how to detect particles. So now I have, essentially, a cyborg. Now we have a chip that can go out, collect the air, analyze the air, and tell you what's in it."
Smarr said he's been noticing the power requirements of something like the exascale: "It takes a million times as much energy to run that exascale as the three of our brains are currently using.... We have to embrace evolutionary design, biological computers - because they're just vastly more energy-efficient."
Agabi agreed, adding that most of the energy in an electrical system is spent moving electrons away: "A single neuron is like a GPU: it does everything contained within itself."
"One of our engineers was doing a calculation on this," said Agabi. "To power our devices for 200 years - engineers like to do this kind of theoretical calculation - would take a single cube of sugar.
"A biological cell - for the programmers in the room - a biological cell is the ultimate recursive function."
Mark Dankberg, Chair and CEO, Viasat
Hosted by Mark Anderson, Chair, Future in Review
On the topic of the future of satellite-based communication, Viasat CEO Mark Dankberg told host Mark Anderson and the FiRe audience that there are currently 4 billion people without internet connectivity, for either geographical or economic reasons. Satellites could close that gap, but it's not a given that they will. The market for entertainment has changed recently: where previously entertainment providers had to work through a broadcast or cable company, now services can be delivered directly on the internet with over-the-top communications.
While satellite launch costs have gone down, Dankberg said the real gains have been in broadband power per launch, which has improved by a factor of 100. When the internet was first available on airplanes, there were no latency issues, but there also wasn't enough bandwidth to support the demand once iPhones hit the market. Then Viasat was able to offer a high-throughput solution with much more bandwidth, which, despite having latency issues, was perceived as a much faster product because there was no congestion.
Regardless of locale, video is the dominant driver of bandwidth consumption and will ultimately determine how much demand there is in each market.
Anderson proposed a quick whiteboard exercise to determine whether there's a demand for a satellite constellation network like the one Elon Musk is launching. Dankberg pointed out that the business case isn't determined by the number of satellites, but rather the unit cost of providing bandwidth from each spacecraft. There is plenty of demand for 12,000-14,000 satellites, but the real question is whether the air-time price they need to make each satellite is attractive enough to be competitive in the market in which it's competing.
Productivity gains in the future won't come from launch costs, but rather from launch payload and network improvements. If satellites are not geosynchronous, the orbit selected dramatically affects your ability to deliver bandwidth economically. For example, in a polar orbit, your spacecraft will spend a lot of time in areas with little to no demand. You can improve this by inclining your orbit, but you may not have access to markets in Russia or China. So there are geographical considerations.
There's no way that satellite-based communication will be cost-competitive with terrestrial 5G in a market where the terrestrial provider is motivated to compete, Dankberg added. Satellites can, however, compete with older terrestrial technology and in markets not adequately served by cable.
Susi Snyder, Nobelist and President, ICAN, and Project Lead, PAX Nuclear Weapons Project; and
Hosted by Kimberly Dozier, Global Affairs Analyst, CNN, and Contributing Writer, The Daily Beast
When you're dealing with intractable problems, change the way you talk about them. - Susi Snyder
Most people's biggest mistake is that they "preload" problems. - Berit Anderson
I don't want to contribute to disinformation, so I do not let myself retweet something unless I've read it. Every time I break that rule, I pay. - Kimberly Dozier
L-R: Kimberly Dozier, Berit Anderson, and Susi Snyder
Moderator Kimberly Dozier opened this session to applause and a few shouts when she said: "This is a conversation I've really been looking forward to - and, may I point out, three women on one panel." She then asked panelists Susi Snyder and Berit Anderson to each give an example of a problem they've tackled. What changes have they managed to bring about, and what did they use to do that?
Nobelist Snyder led with appreciation and humor: "I love being in rooms surrounded by brilliant people, because it's brilliant people who change the world. And it's also a whole lot of fun people. That helps - so don't get too full of yourselves."
She said she'd noticed the question over the past few days of addressing the TCP/IP problem, in which you have to build a certain level of trust. "I work to get rid of nuclear weapons, which is technology from the '70s, from the '80s, and bases itself on a certain level of trust in systems that are no longer present. So I'm listening to a fantastic CTO Challenge, I'm listening to the way this room is changing these perceptions, and I'm thinking, 'I've got some cool stuff that I'm going to take away from here,' and here's what I'd like to offer to you: When you're dealing with intractable problems, change the way you talk about them."
Instead of talking about nuclear weapons as "tools of strategic stability," Snyder said they talk about them "as what they are: big, dumb, clumsy bombs that are designed to indiscriminately mass-murder people." With that switched perspective, global buy-in for prohibiting them becomes possible. "We did it with land mines, we did it with cluster bombs.... we can do it with nukes," she said. She suggested that the CTO Challenge team try changing how we talk about the internet.
Of 197 countries, there are currently 79 signatories committed to enacting national legislation against nuclear weapons. "There's also a lot of pushback," Snyder said. "Only nine countries have nuclear bombs, and they don't want to give up their big toys." In the US alone, $100,000 per minute is spent on nuclear weapons. That said, she's convinced that the more they're referred to as "not a symbol, but a bomb that's designed to kill people," and the more countries that sign on, the closer we are to making that change.
Dozier introduced Berit Anderson, describing her as having "tackled and told one of the most groundbreaking stories about disinformation, which is one of the things that's ripping apart communities and nations right now." She asked Anderson how she did it.
Anderson said that when she was with Scout.ai, which was instrumental in uncovering the 2016 election disinformation campaign, they first decided what to tackle, in a few steps. First you find out about a problem and do a certain amount of research to make sure you get it. Scout "combined near-future science fiction and investigative reporting to help people understand the social implications of technology." They chose near-future science fiction because "it allows you to create a vision of what could be without normal mental blocks. If you drop that frame and just focus on what is the optimal outcome for the situation, you can then start to work backwards and figure out what steps you need to get there and how. What are the road blocks?" Most people's biggest mistake is that they "preload" problems, setting their sights short.
It was through research and conversations that they figured out Cambridge Analytica and Russia were using social media to game the outcome of the US election, and that they'd done the same with Brexit and planned to do it in elections across Europe, Africa, Asia, South America ... "and in fact were very successful in many of those countries." Scout saw the importance of the story and decided to use the same tools they were using. Using social media, they test-targeted responses of different groups, then used their findings to push a "real" message out.
Anderson then arranged to give talks to audiences of influencers, especially in Europe, including politicians and policymakers - because of their sensitivity to privacy issues - about the kinds of solutions they could put into place to prevent the same from happening in their countries. From a whole different angle, she also talked to developers at conferences in Europe, since workers are one of main means of leveraging change inside big tech companies, and these companies don't want to lose their expensive talent. And - for the most part - "the people who actually work for these companies want to be 'forces for good'; no one wants to work for a company that they think is ruining the world."
All that led to a much greater awareness across Europe and helped put into place thoughtful implementations for privacy of user data. Europe led the charge; the US is now catching up.
Dozier said sociologists studying social media have reported that using emotional wording doubles the likelihood of a post being read or shared. As a reporter, especially given numerous tours in war zones, she's always taken care to moderate her reactions: "Everyone's expecting me to be upset and to be unreasonable." She asked Anderson and Snyder if they find it useful to either use or avoid emotional language in their work.
Anderson's background is also in journalism, she said; so prior to Scout she tended to err on the side of objectivity. But she learned from the propaganda experience that it is helpful when talking with audiences to make a story personal; it allows for better connection with the audience and better illustrates the story.
Snyder said that as a woman working in defense and security, the presumption of emotionality can create challenges. "But passion can be a very powerful motivation.... It helps you in some ways get over some of the roadblocks you set for yourself in trying to identify your end goal."
"I know for a fact that I will live to see the end of nuclear weapons. I know that," said Snyder - which makes it easier to go to the evidence rather than being emotional. That said, she thinks a combination of evidence, enthusiasm, and occasional emotion is a good balance: "We work with survivors all the time - of Hiroshima and Nagasaki, but also survivors of the 2,000 nuclear tests that have gone off, people who have been physically affected who are willing to tell their story. It's a very traumatic story."
On the difference between manipulation and messaging, Anderson said she sees a lot of manipulation in her work; she defines it as "trying to get someone to do something that is actually against their own interests." But for effective change-making, you need to identify the key stakeholders and their interests, and figure out how to design a system or solution that really is in their interest.
Snyder: That's why we've been successful within our campaign: we look at data, at who's willing to lean a little more toward [our] direction. Getting to those people and serving their best interests can be very challenging with entrenched mindsets ... a culture where either people don't think about nuclear weapons or they figure it's in good hands.... Also, what we're proposing is not idealistic; we're not saying we're going to end war, or change the world dramatically. We're just going to take one thing and shift our perception on it, and take that one element out.
Dozier: We've talked about two things a lot at FiRe: disinformation [and] ... this battle between communities, this growing incivility. How would you tackle that?
Anderson: On both sides of the line, I see the propaganda that's going on to try to make people be a little more angry at each other.... The biggest barrier is just the personal reactive instinct that humans tend to have. When you start to try to build coalitions and bring in unlikely allies, just coming in with a sense of openness about their perspective and their ideas is probably the most important tool.... [At Scout we would] invite people that might not have necessarily been "friends on Facebook."
From the audience, FiReFellow Wilfred Audley asked Snyder about the challenges of enforcement, including which would be first to get rid of nuclear weapons, Russia or the US?
Snyder answered, in part: "Remember, in 1984 there were 70,000 nuclear bombs in the world. There are 14,000 now.... Russia and the US are two countries that have done an amazing job in bringing those numbers down with one another. And they've kicked off their most intensive arms control talks when they've been at their worst tension. The Cuban Missile Crisis led to a global nonproliferation treaty," with both countries afraid the weapons would go to everyone.
Dozier: So perhaps the breakdown of the INF Treaty could lead to a new surge in denuclearization; we can only hope.
Dozier saved her promised climate-change question for last: "How would you apply some of the lessons you've learned to getting the message out on that?"
Anderson: The good news is, a lot of people are working on it, and there are a lot who care deeply about it. Project Drawdown, for example, has spent a ton of time researching and figuring out which solutions are most important, which parts of industry are most affecting climate change.... [It makes sense to] focus all of our energy on one of those things at a time, or let's create a group that's focused only on advancing that one issue, another that's focusing only that [other] issue.... We've seen a lot of those companies that are a part of that community here this week onstage. The one thing that's been missing from the climate movement is coordinated strategic messaging.... [I]nfighting has been a major disservice to the climate-crisis movement. We're now seeing a much more cohesive strategy, which is, "Yes, we need everything," which cannot come quickly enough.
Snyder agrees that disparate messaging is a problem: "... [W]hat we need is a combination of policy - it's industry, and it's those who are willing to lead the change. This week has been magnificent, because this is a room full of people who are willing to lead the change and demonstrate that it is possible. Because as people see that it is possible, they will get on board.... I love the fact that so many of these people are making it happen."
Dozier recounted a conversation she had with another FiRe attendee she rideshared with on the way to the lodge. To Susi's earlier point, they were talking about climate change, and the other person had said: "We've learned in certain parts of America that don't believe that climate change is real, we just use different words. We use words like resiliency. You emphasize saving money." "Somehow the word's got to get out there that grid parity is nearly here," Dozier added, "and that could be good for America in its competition against China."
"The other thing that I wanted to offer that I always try to do in terms of disinformation," Dozier added: "I don't want to contribute to the disinformation, so I do not let myself retweet something unless I've read it. Every time I break that rule, I pay. I also am really careful not to amplify hate, even if I totally agree with someone's snarky, contemptuous post.
"With that, I thank you both for your solutions and creative ways of solving these problems, and thank FiRe for the opportunity to have this conversation."
The Lodge at Torrey Pines in La Jolla, CA
(L-R in photo) Front Row: Dakota Carter, Evan Anderson, David Sousa, Evan Roche. Middle Row: Mark Anderson, Berit Anderson, Sharon Anderson Morris, Sally Anderson. Back Row: Max Navas, Anne Vertovec, Wilfred Audley, David Morris, Sebastian Damm, Denyse Hudson, Shallon Jordan, Lee Hall, Scott Schramke. Not pictured: Kris Krg, Karen Roche.
© 2019 Sally L. Anderson
Photos © 2019 by Kris Krg/SNS Future in Review Conference Corp. unless otherwise noted. For information about use and purchase of KK's photos, please email firstname.lastname@example.org.
See the full FiRe Gallery sets here, organized by year - and search by keyword to quickly find yourself, others, and topics of interest: http://gallery.futureinreview.com/.
Your comments are always welcome.
Mark R. Anderson
To arrange for a speech or consultation by Mark Anderson on subjects in technology and economics, or to schedule a strategic review of your company, email email@example.com.
We also welcome your thoughts about topics you would like to suggest for future coverage in the SNS Global Report.
For inquiries about Partnership or Sponsorship Opportunities and/or SNS Events, please contact Berit Anderson, SNS Programs Director, at firstname.lastname@example.org.
Please visit www.stratnews.com/insideSNS for:
We look forward to returning to
Join us for four days of industry-changing conversations, problem-solving with global leaders, lively debates, and social events. And you'll connect with the world's most strategic thinkers and doers on tech, business, climate, and the global economy.
Register Now for the best price!
* On November 21-22, Mark will be speaking at the Keiretsu Family Office Network events in San Francisco. * In early January, he will be speaking at the TAG Annual Predictions Event in Bellingham, WA. * And October 6-9, 2020, he will be at the 18th annual Future in Review conference at The Lodge at Torrey Pines in La Jolla, CA.
"Strategic News Service," "SNS," "Future in Review," "FiRe," "INVNT/IP," and "SNS Project Inkwell" are all registered service marks of Strategic News Service LLC.