☀️🔮 Understanding Chinese tech giants; the current limits of our AI paradigm; tribal identity & future politics; genes, eels & cocaine++ #171
Dept of the near future
🚩 Amy Chua: The future of politics is tribal identity. EXCELLENT: “tribalism remains a powerful force everywhere… it has begun to tear at the fabric of liberal democracies.” I'm not sure of her recommendation, which is to re-up the role of national identity to get disparate tribes to cohere. My observation: cities may play an increasingly large role in group and identity leadership. And the impact of technologies, such as machine learning, make it easier to render explicit tribal affiliations. Digital networks and increased mobility make it easier to “find your niche”. So it could be that the cost of understanding which tribe is yours and finding it (wherever it may be), would decline. Equally, the physical realities of adaptation mean that conflicting tribes may need to live and share the same space for long periods of time. I’m curious about watching this trend.
💊 What is wrong with the current approach to AI and how to fix it. We’re trapped by Turing’s definition, rather than tackling the real issue which is common sense and Moravec’s Paradox, argues Filip Piekniewski. (THOUGHT PROVOKING.)
⚔️ Inside the Tencent-Alibaba battle for domination in China. Different cultures, different strategies. “The most obvious [battle] is in their respective investment strategies. The two have been such assiduous dealmakers that [analysts count] 280 Tencent deals over the past three years and 174 for Alibaba.”
🍃 Yay! Good news from the Bloomberg Energy Outlook, our go-to-source on the energy transition. By 2050, renewables supply 87% of electricity in Europe, 55% in the U.S., 62% in China and 75% in India. Fossil fuels are facing unprecedented challenges as the benchmark levelized cost of renewables trends down often at double-digit rates and grid and local battery capacity ramps up to handle supply-demand smoothing.
🏛️ As automation increases, “architecture is losing places where you interact with people” and the nature of buildings needs to change.
Dept of China
🇨🇳 Michael Moritz, one of the most successful venture capitalists of all time, writes that China is winning the global tech race.
Colleagues who come back from trips to China talk, in eye-watering terms of the vision of teams, the breakneck pace of development and the staggering rate of growth. For example, Steve LeVine argues that “Chinese retail is winning the future” and is streets ahead of what we see in the UK & US.
To give you another flavour of this, I recommend watching this talk given by Alan Yui, VP and Chief Data Scientist at Ant Financial on my stage at the CogX event this year. Ant Financial is the world’s most valuable fintech and the fifth most highly valued financial service institution in the world today. (Also recommend this solid profile of Ant Financial in Bloomberg.)
Two things to ponder:
Ant Financial has built a financial services firm with AI at its core. The result is excellent products, better customer understanding and a remarkable innovation rate.
Can you imagine your incumbent bank coming close to this level of vision? Will be interesting if any European or American challengers can achieve the same.
Elsewhere:
Great infographic on where China’s giants are placing their AI bets.
Chinese firm, CATL, built up the biggest lithium-ion manufacturing facilities in the world.
BTW, I haven’t managed to get out to China in the past few years, and I am due for a trip. If you can help with introductions to smart founders, companies that are taking off and great investors, I’d be grateful.
Dept of artificial intelligence
This is the second week where we’re going to review the Cutting Edge stream, which I curated, of the CogX Festival of AI. Recommended is Marija Gavrilov’s summary of the day, which also links to all of the brilliant talks.
If you only have time to watch a few items, I’d like to recommend the following:
Haiyan Zhang on how she used design thinking and machine learning to help a young woman with early-onset Parkinson’s.
Shiva Amiri on giving human scientists superpowers to help them make discoveries beyond the realm of human intuition.
Dave Palmer explores how digital criminals will make use of new AI techniques to run amok (and what we might be able to do about it.)
Elsewhere:
Google Deepmind has filed a range of patents which AI researcher, Ben Goertzel, terms “bogus” and “all or nearly all of which certainly are not their original inventions.”
More startups and researchers are experimenting with approaches other than deep learning to advance AI. Good, old-fashioned AI approaches are gaining currency, as to be expected… deep learning has picked off many of the low hanging fruit. (For a good view of one approach combining deep learning with semantic networks, watch this 10-minute presentation by Marta Garnelo.)
Some suggestions for policy interventions in the case of robot-driven unemployment. (US market context. Some interesting ideas here, although as we discussed last week, the US labour market is so tight right now, this sort of technological unemployment seems far away.)
Adobe will use machine learning in Photoshop to detect tampered images.
An AI system which predicts when hospital patients might die may improve care pathways.
Profile of Corti.AI which uses machine learning to help 911 operators identify stroke and heart attacks faster.
Korea’s fledgeling KakaoBank can handle 80% of its customer service queries via a smart chatbot. (KakaoBank, a service of the dominant local messaging service, is no slouch. It signed up 5.5m accounts in its first 9 months.)
Dept of custom research: Facebook & data use
Introducing our first custom research. We investigated how well Facebook users understood Facebook’s approach to user data. This piece of research was made possible by London-based research firm Qriously and the several hundred readers who took the time to share Exponential View with their friends in the past four weeks. (Thanks!)
In a nutshell, a large majority of US consumers surveyed, wrongly believe Facebook sells their data to third-parties. This, to the benefit of Facebook, does not significantly change their on-platform behaviour. You can read the summary with detailed data here.
I’m open to hearing about things you are curious about you’d like us to explore. Drop me a note.
Short morsels to appear smart at dinner parties
👁️ In 1986, European car maker, Daimler, successfully built a roadworthy autonomous vehicle. Great story with a video.
💯 Ellen Pao: Why “involuntary celibates” (a fringe group, also known as “incels”) thrive at tech companies.
The man who was fired by a machine.
The post-literate: how and why reading is declining in America. 💔
Understanding the massive scale of human sacrifice at Tenochtitlan. (Fascinating that such bloodletting played a key role in helping "define & strengthen group identity"; relate it to the Amy Chua piece above.)
Nice overview of the current state of quantum computing… very slowly getting somewhere.
💥 The 'omnigenic' model of traits: do all genes affect every complex characteristic?
European eels are suffering muscle damage from cocaine.
🌐 Celebrating 20-years of network science. One of my favourite 'new' disciplines, it's had a massive effect on understanding all types of systems at different scales.
End note
Earlier this week I spoke at the Ellen MacArthur Foundation annual summit on the Circular Economy. The circular economy promotes keeping resources in use in the economy as long as possible, before reclaiming components and materials at the end of a products life.
Waste management and reclamation are not generally areas where I have any special expertise. One thing I appreciated, especially compared to a traditional tech conference, was an understanding that innovation can (and, in many cases, should) take place within society and existing social institutions. The enormous opportunity presented by the circular economy -- which will include commercial, financial and social gain -- will probably be unleashed with systemic solutions rather than founders flying solo.
Summary: an intriguing area where I expect we’ll see more investment & huge startup opportunities in the near future.
Next week, I’m working in Iceland and then away camping with one of my kids, so Exponential View will be shorter than usual.
I’ll be back!
Azeem 😎
P.S. Enjoy this? Take a moment to share it on Twitter or LinkedIn or forward to a friend by email.