🔮 Climate tipping points; China’s innovation; open AI; quantum brains & error messages++ #395
Not every climate tipping point is a nightmare.
Hi, I’m Azeem Azhar. I convene Exponential View to help us understand how our societies and political economy will change under the force of rapidly accelerating technologies.
In today’s issue:
Good tipping points
Is China shooting its innovation?
The near future
Not every climate tipping point is a nightmare. Clean energy and electric cars are hitting tipping points for global mass adoption, says Tom Randall. Eighty-seven countries are witnessing tipping points in renewable energy, electric vehicles and grid-scale storage. Heat pumps, which will be critical for domestic warmth, have already replaced a fifth of the boilers in Europe.
One example of a huge grid-scale battery: CATL is building a 1.4 GWh battery in Nevada for solar storage (enough to power more than 50,000 American homes for a day.)
Why this matters: Shifting the narrative around the clean transition is critical to driving more adoption and greater investment. Here is the story I tell: the clean transition is happening rapidly, and it is getting cheaper and cheaper. There might be the occasional bump in the road as supply wobbles and commodity prices rise, but the overall trend is ineluctable.
A Chinese liquor business is now worth more than Tencent as President Xi’s drastic savaging of the tech industry takes hold. The big firms, which drove commercial and technical innovation, and software arenas, are under deep scrutiny as the party moves to limit their power. As Francis Fukuyama writes, Xi’s weakening of the institutions, collective decision-making and decentralisation that made modern China powerful and rich(er) are already starting to hurt: “This concentration of authority in one man has … led to poor decision making.” (Btw, this is a complex picture: China has also taken a lead in permissive autonomous vehicle regulation and is strengthening its domestic IP protection frameworks.)
Why this matters: The Great Leap Forward didn’t work. Nor did the seven decades of the best efforts of Gosplan. Modern economies are too complex for the fickle interpretation of personal diktat. Rather they demand more creative and risky decisions, better handling of uncertainty, and rapid feedback from free-thinking agents in society (be they firms, finance consumers or citizens). The new China may kill its golden geese.
London-based Stability AI, the firm behind Stable Diffusion, raised a $100m seed round at a billion-dollar valuation. Stability runs a training cluster of 4,000 A100 Nvidia GPUs, making it one of the largest supercomputers in the world. The firm is pursuing an open-source strategy, like Red Hat Linux, rather than having closely-guarded models like those of OpenAI or Google. (Check out this post by the firm’s CIO.) Open-source could go some way to broadening access of these technologies. Stability’s models even run on laptops, as I pointed out last month. But there are the huge attendant risks of such models (discriminatory and biased outputs, for example) and it isn’t clear to me that open-source has good governance models for delivering ethics by design. Stability’s licences do have some rails, in so much as the end-user of the model takes responsibility for its legal and ethical use. Traditional corporate orgs, such as Salesforce or Microsoft, have found ways to incorporate ethical frames into their systems.
Why this matters: Turn on your tap. Switch on your lights. Check the ECG on your wrist. The democratisation of technology is usually a good thing. However, there are well-understood risks with large-scale AI models, and I’ll be curious to see how Stability finds novel ways to address them.
Weekly Commentary: Can protocols beat platforms?🔒
In the Weekly Commentary, I look at the prospects of protocols supplanting platforms in social media.
The Commentary will be sent out to paying members of Exponential View. Get a membership for $10/mo to receive my essays every Monday.
Dept of our climate future
In every Sunday edition, we track key metrics that tell us a little about our shared climate future. Our member, Marshall Kirkpatrick, takes the time to curate a view of our current climate status in this segment every week.
One thing to know about our climate future this week: Things are getting worse less quickly; let’s keep turning this ship around! According to a new report from the International Energy Agency published this week, global emissions of CO2 will rise less than 1% this year. Growth is estimated to be 300 million tons to 33.8 billion, down significantly from the growth of almost 2,000 million tons of CO2 added in 2021. The IEA said credit was largely due to record deployment of renewables and EVs. The IEA’s coal estimates for the year may be inaccurately high, too, says Ember coal analyst Dave Jones. According to IEA’s Roadmap to Net Zero by 2050 though, that 33.8 billion tons of annual CO2 emission needs to not only stop growing, it needs to be cut to 30.2 billion by 2025 and then 21.1 billion by 2030. Azeem’s written about the “negative cost of net zero,” or how transition is going to more than pay for itself, if we can execute on that transition. May each move in the right direction be acknowledged, then. As Adrienne Maree Brown advises in her book Holding Change, “Change comes from cumulative shifts. Reflect to groups how they are accumulating change…or not. Reflect to groups what they are practicing - both in and out of alignment with their values.” Let’s keep going!
In other climate news:
Big wind plans: Cumulative targets for new offshore wind generation capacity planned to be built in the US have been revised up this year by 58% to 77.4 GW, according to a new report from the Business Network for Offshore Wind. That’s a lot: current US solar installed capacity as of this Spring was 121 GW and the world’s largest single offshore wind project under construction, the UK’s Dogger Bank, is at 3.6 GW. That means the US now has more than 20X more offshore wind planned than Dogger Bank.
Circular solar: Australian researchers have published what they say is a new method to recycle all but 2%-3% of the mass of a solar panel, at scale, sustainably and affordably. In February, German researchers said they’d devised a method to produce the cells in solar panels from 100% recycled materials.
Billions from South Africa: South Africa has approved an $8.5B energy transition investment plan to move away from coal, though concerns remain about the program’s grants vs further indebtedness and weak follow-through from rich countries that have promised to support programs like this. Separately, South Africa announced this week plans to build a $4.6B green ammonia plant, which will combine solar power and desalinated water from a salt manufacturer to produce fuel to decarbonise the shipping industry.
Tesla is approaching 60m cumulative miles of experience using its “Full Self Driving” mode.
Lenovo is the top PC manufacturer, shipping 16.9m units last quarter. Of the top five, only Apple saw its market share increase (by a whopping 40%).
Post-Covid, Californian VCs invested in more broadly distributed startups. Researchers noticed an increase in geographic diffusion of up to 51%.
The major tech firms have laid off thousands of workers this year.
Short morsels to appear smart while AI recreates history
💊 Absurdly high drug prices in the US don’t come from R&D costs.
🖥 Zero-trust is losing credibility as a cybersecurity paradigm.
🧠 Our brains might use quantum computation.
🐣 Challenging the narrative: the defining trend of our time is regionalisation, not globalisation.
👻 Ali Abdaal tweeted a list of productivity hacks and got over 1 million impressions. The catch: it was entirely written by an AI.
🙀 Do not read this article on high reactance. (Now why would I say that? Better read the article.)
I was in Chicago last week giving a talk on The Exponential Age. It was a great event, probably the single best stage I’ve spoken on and a couple of thousand attendees. (You can see some pics here.)
The trip gave me a chance to catch up on some reading, listening and watching. Three recommendations:
This podcast with EV reader, Michael Liebreich, and Silvia Madedu on decarbonising industrial heat through electrification
Chris Miller’s exceptional book on the story of semiconductors, Chip War
And Disney’s Andor, a Star Wars prequel.
By the way, if you want me to talk at your event (or brief your board or executive team) on the Exponential Age, drop me a line and I can connect you to my agents.
P.S. I am speaking at Wired Impact 2022 in London on November 2nd. Other speakers include BioNTech’s Uğur Şahin, Gaia Vince and Kim Stanley Robinson. The first 20 EV members to sign up will receive complimentary tickets — members of EV will receive more information on this separately tomorrow. For everyone else, we’ve got a 25% discount.
P.P.S. Our own Chantal Smith is hosting an EV meet-up in Berlin on Sunday, Nov 6, 4.30pm. Register here.
What you’re up to – notes from EV readers
Faridah Abdulrazaq is looking for candidates for the BloombergNEF Pioneers Competition.
Packy McCormick is featuring Jocelynn Pearl and Elliot Hershberg on DeSci in his Not Boring newsletter.
Gianni Giacomelli published two videos introducing superminds.
Quinn Emmett published an Important, Not Important podcast episode with John Semmelhack on how to electrify homes.
Eduardo Plastino and his team at Cognizant have published an article about what business executives are missing about the metaverse.
To share your projects and updates, fill out your details here. Because of space constraints, we prioritise updates from paying members and startups I have invested in. (You can become the former by subscribing, if you have not already, and the latter by getting an intro to me via a trusted contact.)