🔮 Building virtual worlds; Meta drops; CAR-T, crypto in North Korea and infinite bubbles++ #358

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.


Today’s edition has been supported by our knowledge partner, McKinsey & Company.

The net-zero transition. Governments and companies worldwide are pledging to achieve net-zero emissions of greenhouse gases. What would it take to fulfill that ambition? From the six key features of the economic transformation to actions stakeholders can take, you don’t want to miss this important research.


🎙This week on the podcast


I spoke to Jonathan Wolf, CEO of healthtech startup ZOE, about how his company uses cutting-edge microbiome science and machine learning to give users tailored nutritional advice. Research on the gut microbiome is relatively young, but ZOE’s studies have shown the outsize impact it can have on our health. In this podcast, Jonathan talks me through how he’s blitzscaling science research, and what ZOE might do next.

You can listen to our conversation here or read a full transcript. Notes are available for members of Exponential View.

The near future


♾️ Mark’s Meta misery
Shares in Meta Platforms fell 26.4% on Thursday after the company said Facebook's daily active user numbers dropped for the first time. The plunge erased more than $230bn from the company's market value, a record single-day drop for any US company. Meta said Apple's iOS privacy changes would decrease sales this year by around $10bn.

Back in March 2018, I wrote that Facebook “is dependent on monetising its users through advertising, and the value of those adverts is dependent on garnering as much of the consumer attention in the world as possible, together with figuring out how to get as much consumer data as possible.” The privacy changes announced this week show exactly how valuable that data is to Facebook/Meta – and the severe effects of limiting its access to that data.

Like its success, many of Facebook’s problems stem from its founder. To understand how the firm can reverse its character of decline and malign to something better, we would do well to look to Microsoft as a precedent. Back to my 2018 essay:

At the peak of [Microsoft’s] power, the firm was the scourge of trust-busters for several years because of the dominance of Windows. While it was the Internet, a fundamentally different technology platform, which tempered Microsoft’s power, it is valuable to understand how long it took to adjust company culture and execution.

Fourteen years. More than a decade, and it involved two changes in CEO, first from Gates to Ballmer and then from Ballmer to Nadella. It remained profitable in that time, but its share price was flat and ennui was its commercial stock-in-trade.

Since Nadella’s appointment, Microsoft has been on a run with its share price more the doubling and its pivot to a cloud-based business model (and different operating practices) successfully completed. Changing values, culture and business model takes an awfully long time.

In other Meta analysis: Megan McArdle on how Facebook is learning the important lesson it dealt us.

👾 Rhyming the metaverse
The metaverse has been with us a long time. Anyone who played in early text-based adventures, like Dungeon, entered into an alternate reality. Sure it wasn’t immersive, the interface was clunky, and there were no graphics – but it was still an alternate reality in its time. We are here today with the promise of immersive, high-definition 3D worlds because of this exponential march of evolving technology.

Raph Kosters, a game designer behind games like Ultima Online, makes that point in a recent talk: that we’ve been here before and we can learn from our experiences (by which I mean our experiences of building 3D worlds, not our experiences in the metaverse.)

Raph disenchants several notions about virtual environments in one key slide:

But with analysts forecasting real money to be made, it’s not going to dampen businesses’ enthusiasm towards this new platform. Morgan Stanley estimates the metaverse might become an $8 trillion market. Mark Zuckerberg is obviously pinning his fortunes on it. So too is Satya Nadella who, in an interview worth reading, explains why he sees gaming as the next platform and how Microsoft will play there (we wrote about this in EV#356).

🧰 Autodidactic autocoders

According to its recent performance in a series of competitive programming games, AlphaCode is better than 54% of human programmers. This new coding engine from DeepMind has demonstrated that it is better at coding the 99.8% of all humans – proven through its trials on the CodeForces platform.

AlphaCode could become a powerful tool for anyone wanting to develop software; a helpful wingman, so to speak. But this isn’t just about efficiency. Several experiments in human-machine teaming suggest it may result in better, more effective coding solutions. This academic paper, Will Humans-in-The-Loop Become Borgs? Merits and Pitfalls of Working with AI, suggested the nuanced benefits of humans and machines working in tandem. AI-human pairs have better judgement than either a person or a program working on their own.

The one drawback is: humans can put too much trust in the machine, occasionally subordinating their critical thinking to its decisions. This speaks to some of the design considerations. If humans are willing to trust machines, even limit their judgment when using automated systems, designers must be exceptionally careful about which cues they use, allowing the human user to maintain agency over their actions.

In other news, an OpenAI network was able to solve some high school maths olympiad problems, demonstrating “that deep learning models are capable of non-trivial mathematical reasoning when interacting with a formal system.” This means AI can successfully identify mathematical proofs, even without being given the formal system that humans use to reach the same conclusions. Although the experiment states that its machine-solver can only crack “some” problems, it is still especially impressive - there is an infinite amount of action space, which is the number of decisions available to the machine/student/learner. As this AI language model develops, it may eventually be able to match best-student performances for the same problems.

Sunday commentary: Pudding and meat


What is the relationship between the academy and economic outcomes? A recent paper in the peer-reviewed journal, The Journal of Education and Learning, explores this question through the lens of literacy rates in Western Europe in the 16th and 17th centuries.

The Four Elements: Air. Joachim Beuckelaer, 1570.

That’s an interesting period: it encompassed the Dutch Golden Age and signified the start of England’s transition from a stinky, poor backwater locale to an imperial titan. Eskelson points out that between 1500 and 1700, literacy rates in the Netherlands and England

experienced divergent growth from the rest of Western Europe due to the preceding establishment of more inclusive and efficient institutions. Such institutions allowed for a society with structures and rules that better integrated markets, led to the growth of towns and cities… and created better incentives for families to invest in formal education for their children.

In other words, the development of economic institutions preceded the development of educational ones. But more than that, the societal structures that favoured more transparent rules and market mechanisms encouraged people to invest in education.


The rest of the Sunday commentary essay is open for members of Exponential View.

Dept of our climate future


3333 days remain before we hit the 450ppm mark, at our current rate.

In every Sunday edition, we track key metrics that tell us a little about our shared climate future. Read here to find out what these metrics mean and why I consider them relevant.

We have invited EV member Marshall Kirkpatrick to curate stories about our climate future in this section every week. “This week’s stories”, Marshall says, “are a great illustration of the momentum-building underway around the world, our expanded understanding of who and what matters, of taking responsibility for the legacies of the past, and investing in alternatives quickly. May developments like these help build the foundation for much, much more.”

  • Global investment in energy transition totaled $755 billion in 2021 – a new record, and a 21% increase from 2020 with almost half the investment occurring in Asia, according to the latest annual report by BloombergNEF. Additionally, climate tech companies raised $165 billion in equity financing last year to scale existing solutions. The report notes that investments in energy efficiency and grid improvements weren’t included, making this a conservative number for the year’s estimate, but helpful for understanding the scale of just under $1 trillion. China represented $266bn of this investment, up an amazing 60% year over year. The US was the 2nd largest single country at $114bn, up 17% year over year. Is it enough? Far from it – the report estimates that in order to get on track for net zero emissions goals, this annual investment must triple between now and 2025, then double again between 2025 and 2030.
  • Open Science publisher PLOS has launched its first journal dedicated solely to climate science. Included in its first papers published are several opportunities for significant emissions reductions, as well as a great article on collaboration between Indigenous and Earth sciences.
  • The US federal government will award $1.15bn in funding for 26 states to plug abandoned oil and gas wells. The number of “orphaned wells” leaking pollution and greenhouse gasses around the country is unknown, but estimates range widely – from tens of thousands to several million wells.
  • One thing COVID taught us is that infrastructure can be built quickly when the will is there. While too much of the world stopped at plexi-glass shields in retail shops, the Philippines offer some great inspiration. After COVID shut down public transportation, the nation of 110 million built 500KM in new bike paths in under a year. This equates to building more than 4500 American football field lengths in less than 365 days.

Short morsels to appear smart while AI takes over coding


🥡 Competing delivery startups are on the outs -  the business of supplying customers with instant groceries is experiencing heavy losses. The year 2001 is calling and wants Kozmo.com back.

👏 Scientific conferences going online has greatly increased diversity and inclusion: female attendance grew by 253%, and gender queer scientists by a rousing 700%.

🧪 CAR-T cell cancer therapy, an experimental treatment turning immune cells into tumour-killing cells, is shown to hold back the disease for more than a decade.

🤔 Florida art collector Desiree Casoni is figuring out a way to hang up her NFTs on the wall.

🔬 Multimodal omics are an exciting new technology that could transform personalised medicine by opening up new avenues of biological understanding.

😄 Daniel Solis designed an AI that creates fictional birds based on old book illustrations. Are these the types of beings that could populate the metaverse?

🔥 Independent hacker P4x got slighted by North Korean spies, and took revenge by shutting down the country’s internet.

🌍 Stablecoins could be a major opportunity for Africans to participate in the world economy.

🆒 New, everlasting bubbles made out of water, glycerol and microparticles can stay intact for more than a year.

End note


If you are wondering about the title of today’s sunday commentary, listen no further.

Thanks to EV member Gianni Giacomelli for his help with the Near Future piece on human-machine teaming.

Thanks to EV member Paola Bonomo for sharing the 16th century Italian poem Orlando Furosio in response to the last week’s commentary on crypto and the mirror of Erised.

Cheers,
Azeem


What you’re up to – notes from EV readers

Claire Curry and her team at BNEF published their annual review of the amount spent on scaling the energy transition and climate tech in 2021.

Paul Dowling has launched an incubator DAO to democratise and diversify the startup funding process by providing capital and support. They’re looking for sponsors, investors, experts, mentors and founders. Reach out here.

Adrien Hobt and his team at PCH Innovations published an article about how robotics can be used to scale circularity.

Quinn Emmett discussed vaccine inequity with medical doctor and academic Dr. Madhukar Pai.

David Klenert published a paper about the relationship between industrial robots and jobs.

Tony Pan’s startup Modern Electron has raised $30 million from investors including Bill Gates to decarbonize heat.

Paul Armstrong is organising the TBD - Technology, Behaviour and Data - conference on 31 March. EV members get 33% off - use the code EXPONENTIAL33 at the check-out at this link. The discount is offered on the first-come, first-serve basis.

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.)


Exponential Jobs - Featured roles for our community
Every week, we curate jobs for our community, by our community. Our members are hiring in climate tech, blockchain, space tech, healthcare, future of work, among other areas. Here are the highlights:

40+ other jobs are available here.

If you’re hiring, post to the board. Annual members of Exponential View can post for free; let marija@exponentialview.co know if you’d like to learn more.


How happy are you with today's wondermissive?

Unhappy | Meh | It's good | Great | Extremely happy, will forward

Comments

Sign in or become a Exponential View member to join the conversation.