The Scaling Up Problem – Part II/ the Biggest Technology Wins in the Breakthrough Climate Bill/ Why Is the Human Brain So Efficient?

The Scale Up Problem – Part II. The?report?on scaling up deep tech in Europe that we launched last week stirred some good discussions (together with the?Sifted?article). Among the feedback I received, one stood out. It is the input from a prominent European investor. She adds four key points that we didn’t cover in the report, but that I think are highly relevant and that I want to share.?

While in the Report we covered the role of corporates as partners and investors, we didn’t address the point of corporate as LPs (i.e. the investors in VC funds). It would be in my view a very good complement to their R&D activities (and on average much better for them than a CVC). The regulatory point she raises is one that I am learning more and more to appreciate as a barrier to scaling up, while it could be actually a leverage… And the lack of funding for women-led companies is something that is getting deservedly more attention. But we missed to capture it appropriately, also because it is not only a European issue. Last but not least, the role the European Investment Fund (EIF) could play in getting deep tech funds started is huge…

Here is the direct feedback I received, which adds more color to it:

“…I would add as follows:?

• Revenue is always better than investment so corporate partnerships are key. My view is that corporates would make great fund LPs and that is still far from happening, although there are the first good signs. Examples: Valeo of course, Telefonica, Maersk, NOKIA – many others are needed. Sometimes when corporates invest as LPs they still do not focus on the hardware/science side of things to a large enough degree, rather it's easier to do more digital companies and that seems the preference. An example that is great and I particularly like is Maersk and Einride's partnership, I believe Maersk is also an investor in one of the VCs and in the company.?
• Regulatory barriers – to me one of the main issues with fundraising is that it takes so long to get any kind of regulatory approval (CE mark, MDR, and so on) for a start-up that the risk is too high for investors and the necessary runway is too long.?This is a major issue because, on the one hand, Europe is financing a company, on the other, the regulatory processes are slower than elsewhere and no one is helping these companies properly prepare for the processes. Europe is then surprised that the company prefers to go to FDA, sell in the USA or do clinical trials in Australia or Kazakhstan, and so on. If it chooses to sell on US market then US investors follow – that’s obvious and natural and actually beneficial as they know processes there. Moreover, it seems approval processes in the US are so much more predictable and plannable!?
• Lack of funding for women-led companies that are usually so well aligned with SCGs, the challenges of now and tomorrow (which can only be tackled with significant technology advancement). Many women scientists are unable to raise funding. Only 2% of VC funding goes to women (in Europe, every year) while the alignment with SDGs is much greater for the female founders. Moreover, let’s look at med tech/ biotech. These female founders are tackling the challenges of female health. Yet only 4% of R&D money is spent on female bodies research (R&D, clinical trials) – recently McKinsey said 2%. Yet medical and life science has to do with 50% of the population and that is not getting funded or noticed. This is also science, it is IP based, and it is deep tech. Huge gap, no action.?
• EIF strategy is a major issue for European deep tech, they mainly fund digital funds (and also they do not anchor). How can deep tech funds be born if there are no anchor investors for them – and the only pan-European financial institution whose job it is to do this, focuses pretty much on digital (and talks about the success stories of 15-minute food delivery and scooters)? I suspect it may well be similar to local national fund of funds, very few actually fund deep tech VCs, inter alia because the results are not immediate. Sometimes there are positive examples – Finland – but rare still. These positives should be highlighted!”

Here Are the Biggest Technology Wins in the Breakthrough Climate Bill

The?Inflation Reduction Act of 2022?represents the?largest climate investment in US history?and - having?recently passed?the Senate - appears to be well on its way to becoming law. According to?Leah Stokes, Professor at UCSB: “This is the transformative clean energy and climate rescue package that we’ve been waiting for.” The bill provides “nearly $400B for climate and energy projects.”

The bill provides billions in tax credits for clean energy projects as well as “$60B in incentives for domestic manufacturing of everything from batteries to solar panels to heat pumps” and $27B for clean tech R&D.

News items:

How the New Climate Bill Would Reduce Emissions

A?new study?by the?Repeat Project?predicts that the new climate bill could reduce US greenhouse gas emissions to “roughly 40% below 2005 levels” by 2030.

Why Is the Human Brain So Efficient?

The human brain is “10M times slower than a computer.” So why does it still have “superior flexibility, generalizability, and learning capability” compared to even the most powerful computers and AI models? One reason is “massive parallel processing.” Each neuron in the human brain “collects inputs and sends outputs” to other neurons “on the order of 1,000 on average.” By contrast, a transistor only has “three nodes for input and output altogether.”

For decades, “engineers have taken inspiration from the brain to improve computer design.” The more neuroscientists learn about the brain, the more engineers can use that knowledge “to further improve the architecture and performance of computers.”

News items:

The Science Part of the CHIPS and Science Act

The ?National Science Foundation?(NSF) could receive ?$81B in new funding?under the ?CHIPS and Science Act?- its “largest funding increase” since being founded in 1950.

Deciphering China’s AI Dream

Jeffrey Ding, Postdoctoral Fellow at ?CISAC?and author of the ?ChinAI newsletter, reviews his 2018 ? Deciphering China’s AI Dream Report?and surveys recent changes to China’s AI landscape.

The number of AI startups in China has dropped precipitously since 2017, with only 57 counted by ?IT Juzi?in 2021. However, late round and IPO financing has “rebounded to peak levels,” leading IT Juzi to conclude that “China's AI field has passed through the capital market’s cold winter period, and there is no obvious low-valley period.”

The AI subdomains most favored by financiers include “intelligent finance,” robotics, transportation, and logistics - while “fields such as storage and AI chips” are “relatively weak.”

News items:

Scientists Reanimate Dead Cells in Pigs, a Potential Breakthrough for Organ Transplants

OrganEx?- a new technology ?developed by Yale researchers?- successfully “reanimated” cells in the organs of dead pigs - potentially “confounding conventional wisdom about life and death.”

You Were Asked to Create the Future — This Is What You Said

In Ireland, the ?Creating Our Future?initiative surveyed over 18K people nationwide about “the role that research can play in creating a more sustainable and equitable future.”

Overall, the initiative “reveal[ed] a desire for a holistic, not a technocratic future, a future founded on the values of inclusion, social justice, and environmental sustainability” and organized the findings into 16 themes, such as “Embedding climate action across society” and “Safeguarding public interest and trust in the digital world.”

News items:

I Talked to Leonardo da Vinci

An “interview” with?the?Renaissance man Leonardo da Vinci (based on his notebooks and ?Walter Isaacson’s biography.) Learn Leonardo’s views on higher education and why dissecting a horse didn’t make him a better sculptor.

A.I. Is Not Sentient. Why Do People Say It Is?

Desdemona, the singing robot

Despite ?ex-Googler’s?and ?OpenAI executive’s?claims to the contrary, AI is less sentient than “a worm,” and LLMs are “no more likely to be sentient” than a “rock.” ?Alison Gopnik, Professor at UC Berkeley, said: “We call it “AI,’ but a better name might be ‘extracting statistical patterns from large data sets.’”

So why do LLMs and other current AI models have even the experts fooled? According to ?Andrew Feldman, Founder and CEO at ?Cerebras: “There are lots of dudes in our industry who struggle to tell the difference between science fiction and real life.”

News items:

Tech Talent Wars & the Role of Ethics in Big Tech Success (Long-Term)

An interview with ?Facebook whistleblower?Frances Haugen?on Big Tech ethics and why Facebook “struggles to recruit top talent because of their negative image in the industry.”