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.

👋😍 To level up on my thesis quickly, pick up my book, Exponential. It is now available in paperback. (It was the 19th best-selling title on Amazon this week – out of millions of choices!)

The near future

😮 Ooober
Uber surprised a lot of us and turned cash flow positive in the second quarter. The beat – $308m vs expectations of $109m – is remarkable. So too are other numbers: 1.87bn trips in the quarter, up 22%. This might mean that many of the optimists (and I wasn’t one) were right: Uber could hit the kind of scale that makes its business work. Uber has taken 13 years to get this milestone. Amazon took 14 years before its first profitable (and cashflow positive) year. Can Uber follow in its footsteps? Hubert Horan has been a meticulous analyst of Uber and is not convinced believing “[t]oday, Uber is offering much worse service at much higher prices than the traditional taxi industry that it had ‘disrupted’”.

See also: Amazon’s second act, cloud computing, is on a tear. With more than $120bn in capex since it was created, AWS revenue is nudging $20bn a quarter,(with operating margins around 25%) up from under $5bn a quarter five years ago.

📝 Not worth the paper
More than a quarter of a million international treaties – think of UNESCO or the WTO or NATO – have been signed over the past few hundred years. A meta-analysis of 82 high quality studies suggests that most of them fail to meet the objectives they were designed for. The treaties most likely to meet their goals were the ones which were signed by a small number of parties, were focused on economics and trade of finance, and had suitable enforcement mechanisms in place. The fluffier treaties had little impact and, perhaps, served as reputation-washing for bad actors. As we rethink what economic and technological globalisation means, this analysis may help us think about how to better design systems for international cooperation.

⛱️ Hot act, cool jobs
The climate impact of the weirdly named Inflation Reduction Act will be significant. It should reduce cumulative GHG emissions by about 6.3 gigatons and help the US close about 2/3 of its emissions gap to 2030. (The US emitted about 4.9 gigatons of the bad stuff last year.) So many of these policies exist in systems with feedback loops which are hard to model. We’ve seen electric vehicles and solar power outstrip most expectations, so I wouldn’t be surprised that – if passed and other things being equal – the impact is greater than this. Such shifts will create demand for all sorts of new highly-skilled jobs. One estimate is 9m new jobs in the US alone, about a million a year for the next decade or so.
This level of job creation might be a bottleneck. The UK’s NESTA looks at the need for heat pump installers in the coming years. Britain currently has 3,000, and likely needs 27,000. Achieving such rapid upskilling might mean policy interventions to ensure coherent pathways for reskilling while maintaining the same quality of work.

Sunday commentary: Policy as code

I invited EV subscriber Vass Bednar to share her thinking on the intersection of policy and technology in today’s Sunday commentary. Enjoy, and thank Vass by sharing her essay on Twitter.

Hi, I’m Vass Bednar. I’m the founder of “regs to riches,” a senior fellow at the Centre for International Governance Innovation (CIGI), a Public Policy Forum Fellow and the Executive Director of McMaster University’s MPP in Digital Society Program. I spent two years on Airbnb’s public policy team helping to bring forward regulations for short-term rentals from 2017-2019. Much of my current research and policy advocacy is focused on improving competition law in Canada. In this guest essay, I’m switching gears to offer a perspective on a new role for technologists to better inform public policy design and accelerate implementation.

Uber’s aggressive tactics in the 2010s inappropriately characterised the space between tech companies and governments. The adversarial legacy is characterised by scepticism and dissonance between the two sides. As residual mistrust endures, it hurts the broader policy process by perpetuating a disconnect between policy ideas and their technical viability.

As the terms of policy priorities are debated, technology firms may insist to government officials that technical compliance with a proposed regime is too difficult - if not impossible - to implement. A government currently has no real way to validate such a claim -  though it may occasionally be true - for instance, the initially proposed Online Harms approach in Canada imagined humans would be able to enforce a 24-hour takedown standard. This was going to mean that the government would have to hire thousands of people to monitor the internet, without any clarity on how this was going to happen or how it would be resourced.

In reconciling the demands of technology to the traditions of public policy processes, we have under-explored the need to infuse the expertise of coders into policy design early and often. After a new law is passed, tech firms of all sizes typically have to translate new rules directly to their tech stack. And this is not always possible. That’s why any digital policy needs to be re-articulated in proposed open-source code as part of introducing legislation.

It’s time for a new linguistic layer that expresses the viability of a policy proposal through technical specs.

📧 This is an excerpt. Vass’ essay in its entirety was delivered to your inbox separately. Check your email.

Short morsels to appear smart during the crypto meltdown

📉Rana Mitter on China’s looming demographic disaster. See more in EV#383.

🤖China produced 366,000 industrial robots in 2021, up nearly 68% on the previous year. See also, the AI startup creation boom in China is over. Only 57 AI companies received venture backing last year compared to over a thousand a year between 2015 and 2017. Total value of investments didn’t slow down with RMB4bn ($600m) deployed last year up 53% YonY. (Unsurprising: mature firms need more capital than fresh startups. Full report.)

📱Japan was the world leader in semiconductors in the 1980s. It blew its lead and is racing to catch up.

⛓️Did two brothers use bogus identities to hype up the Solana blockchain?

🤨 Social norms and honesty: “Individuals who hold very strict norms of honesty are more likely to lie to the maximal extent”.

🎖️ The US Space Force replaces annual fitness tests with continuous monitoring. h/t Vishal Gulati

End note

I’m sort of still on holiday so thanks to Vass for stepping in with a thought-provoking commentary this week!

Cheers,
Azeem

P.S. All members with annual subscription are invited to apply to join our awesome community on Slack. Apply here.