🔮 Oh, Facebook! IQ, speeding evolution, slowing science; designing chatbots; Chomsky vs Piaget; clothes rental, exoplanets, the poop catalogue++ #192
Chomsky vs. Piaget
Dept of podcasts 🎧
For so long, we have neglected the workforce of public service. We have not invested in it. We have not helped them get the skills. Imagine you're running any company in the world and you don't have a skilled workforce that's up to date on the trends, of course they're not going to perform.
I had the great pleasure to talk to Lisa Witter, the co-founder of Apolitical, a startup connecting public servants across the globe around knowledge-sharing and best-case policymaking. We discussed the state of public service, equipping government with skills for the digital age, and reviving trust in public institutions through storytelling.
Subscribe and listen to the Exponential View podcast here:
iTunes | Spotify | Stitcher | Overcast | Breaker | Soundcloud
Dept of the near future
🤦 Oh, Facebook! A New York Times investigation suggests that Facebook took a strategy of "delay, deny and deflect" when dealing with the fall out of Russian disinformation on its platform. This went as far as hiring a publicist to do a hatchet job on anti-Facebook groups and link them to George Soros in an anti-semitic twist. Mark Zuckerberg also reportedly ordered Facebook execs to use Android phones after Tim Cook criticized the company. (This is one of the most powerful companies in the Western world, after all it has a substantial role in determining what people read. In the words of EV reader and former President of Estonia, Toomas Hendrik Ilves, is "congenitally ethically challenged. For 15 years the company has been taking and retaking Ethics 101." The problem is that companies do, really, grow from their founding seed crystal, in this case, the peccadillo of a teenage boy. It will take many years for Facebook to fix itself, but it may have reached its peak in its current incarnation and will need to go through a painful period of reincarnation, much as Microsoft did in the late nineties. One piece of evidence: "use on its core platform is down 20%" year on year. A second: Facebook employees are markedly less optimistic and proud to work at the company. People do the work, and if people start to leave the quality of the work will go down. See also this good discussion on whether employees can change the ethics of the firms they work for. Not easily, it seems. )
💸 Christine Lagarde, head of the IMF: the case of digital currency. See also, this Bank of England essay, the seven deadly paradoxes of cryptocurrency.
🔊 Clive Thompson explores the future of voice-based chatbots in this very nuanced essay. (Evidence that bots can help people with depression; the rise of "character leads" who give bots the right 'personality'; and dangers of "creat[ing] objects that inquire after our needs."
💯 Why is science getting fewer results even as more money is applied to research? What is the impact on our future? And what could we do about it? (Excellent read)
🤔 Are we ready for this? Steven Hsu's new genetic test can predict the risk of an IVF embryo having a low IQ. (See also: evidence gathers that humans are evolving faster than ever. And curiously, humans appear to be getting better than ever at playing Tetris. Tight-knit player communities seem to be the catalyst.)
🏳️ Robots, drones & stealth jets: Katrina Manson on the future of warfare. (See also, this sobering overview of the US's greatest security threats: cyberattacks, domestic division, Pakistan, China and more)
Dept of appreciation 💌
The EV newsletter is provided for free. If you are reading it and value it, please make sure that you have shared it to your social networks (Facebook, LinkedIn or Twitter) or by email to friends at least once in the previous four months. Last week, more than three dozen of you took time to do this! Thanks!
Dept of AI and automation
Super profile of Fei-Fei Li, the leading Stanford AI researcher, and her mission to make AI better for humanity.
💭 The debate between Chomsky vs. Piaget has parallels in today's approaches to machine learning. With neo-Chomskyians arguing the real AI systems will need some hard-wiring to enabled rapid learning, and Piagistes arguing for a blank slate model of learning (best thought of today as the massive-data, deep learning coterie.)
🇩🇪 Angela Merkel presented the AI strategy for Germany which includes a €3bn plan, and a focus on healthcare, manufacturing and automotive applications of AI. The company intends to use industrial, rather than consumer data, as the bedrock of German AI investment. This makes sense given Germany's dependence on the automotive industry, and the three-fold attack on that industry from autonomous vehicles, ride-hailing and electric powertrains.
And in China, Xi Jinping led a Politburo 'study session' which emphasized the continued importance of AI in Chinese development and governance goals. Xi argued that:
AI is characterized by multidisciplinary integration and a high degree of complexity. We must strengthen deliberations and decisions, plan matters comprehensively, coordinate and innovate, move ahead steadily, and make strengthening original innovation capability into a focus point
Meanwhile, levels of private investment in Chinese AI are down 30% on last year. Valuations have run ahead of expectations.
Digitisation is coming to Asia's fashion factories. Together with increasingly advanced robotics, this may create "job losses and social unrest". Good interview with Pamela Mar, who I met last week.
Google has released an update of its Open Images dataset containing 9m images with tens of millions of annotations. For many groups, access to data is the gating factor for developing a useful AI application. So, it's all very encouraging. The question I like to ask is whether such things (annotated data) need to be more vigorously available in the public domain through voluntary open-sourcing or other approaches. (See Germany's AI plans above.)
Dept of internet services
Are closed Facebook groups the future? Their membership is up 40% year on year.
Amazon has 10k employees working on Alexa.
🦀 What were the most popular items bought on Alibaba's singles day this year? ($31bn of items were sold!) (Alibaba tracks order levels per home in Beijing, according to this photo.)
Renting rather than buying: the clothes edition. Nice profile of Rent the Runway, which now has 10 million members, who subscribe to rent clothes. Rent the Runway started with high-end brands and is now making its way to more mainstream ones. (See also, Volvo launched a rent-don't-buy scheme, called Care, at $650 per month. It is exceeding expectations.)
WeWork raises $3bn from Softbank. $42bn valuation, annualised revenues approach $2bn, 354k desks globally with occupancy of 84% (!)
LinkedIn's media biz will bring in $2bn in revenue.
👌 See also, a great essay by Rahul Vohra on how to achieve product-market fit.
Short morsels to appear smart at dinner parties
Human history in one chart (BEAUTIFUL!)
💔 Why are young adults having so little sex?
🚌 Thanks to South Korean Internet cafes, Fortnite hits 8.3 million concurrent players. (That represents 0.1% of all humans.)
Exporting social credit: Chinese firm ZTE helps the Venezuelan government introduce a pervasive new citizen ID.
🏠 A fifth of Chinese apartments (50 million) is unoccupied.
Researchers create a promising hybrid photocell that creates both electricity and hydrogen.
Fascinating insights about what worries Americans told through 30 years of letters to an Agony Aunt. (Sex, sexuality and religion top the charts.)
🤩 Quantum biology: the first successful entanglement of bacteria with photons.
A large exoplanet is orbiting Barnard's Star just six light-years away.
🛰️ FCC approves SpaceX's plan to deploy a record number of satellites to orbit.
💩 Global Seed Vault's new cousin, global poop catalogue, created with hopes for medical breakthroughs. See also: do gut bacteria make way to our brains?
End note
We run the risk of a lack of trust around what I am going to refer to as small-t technology.
Small-t technology is technology as defined by the wonderful W. Brian Arthur, "a means to fulfill a human purpose". And technology should be a humanistic, progressive, diverse, inclusive arena which uses the tools of curiosity, observation, cultural interaction, and the scientific method to "fulfill a human purpose." Small-t technology is part of everyday life. It is the smallest tweaks in code, organisation, social behaviour, widgets, practice, communication that feed into the economy and our groups. That, in turn, creates new tensions from which we make more micro-innovations. Small-t technology in this recursive, fractal process that enabled us to build civilisations, reduce workloads and burst through the Malthusian trap. It's rather important to us. And importantly, it is second, or perhaps, first nature.
But small-t technology is often confused with Big-T Tech, or just bigtech.
The real (and exaggerated) misbehaviour, stubbornness and obnoxiousness of many the individuals in the Big-T Tech industry (execs at Facebook, Google, Twitter and beyond, and their firms in toto) might spill over into a general diminution of faith in small-t technology. The probable bursting of the hype bubble around some types of AI in the next 12-18 months won't help.
The trouble is that small-t technology is a necessary but not sufficient requirement to improving life—tackling climate change, improving education and access, supporting refugees and migrants, dealing with the demographic transition, enabling improved forms of democracy, improving welfare and living standards in multitude of ways.
And a turn away from small-t technology's beneficial aspects could be dangerous. True, we've been uncritical of it for too long. And too many still believe that even small-t technology is value-neutral. (This isn't a view EV holds. One view on how technology isn't neutral is in the guest issue of EV edited by Deepmind's Mustafa Suleyman last year.)
So, we need two different strategies: one to tackle the power, where it distorts, of bigtech, and another to critically evaluate, steer and design the direction of small-t technology and the system within which it operates.
Have a great week!
Azeem
💬 P.S. Scroll down for your fellow readers' news and amazing projects. The EV community rocks!
What you are up to
Tucker Max pours his heart out writing about how (illegal) MDMA therapy changed his life.
Congrats to Zavain Dar for promotion to Partner at Lux Capital!
Sean Gourley raises $40m for his firm Primer.ai.
Sheena Matheiken is premiering her film Sea to Shining Sea in London in two weeks. Tickets are disappearing fast.
Kevin Werbach's excellent new book on trust and the blockchain is available from Tuesday.
Keep on sharing your news and projects, email Marija.