Jonatan

Sovereign compute

My simple test whenever I meet anyone in DC making strong claims about China is “Do you know who Deng Xiaoping is?” And the failure rate is way higher than you might think

Only asking for the steel man of why not to take in an angel

Tell me the worst thing an angel investor ever did to you

@FlorianSeibelQS can we beat US to it?

@ben_sdl @SkanskaGroup Maybe! Will dm

@mountbranch @SkanskaGroup 🫡 Do you have a connect by any chance ?

Day 3 of seeing how we can deploy physical AI in Europe. I was at 11 companies in 2 days, getting to speak to the right person is an art, more here:

@CJHandmer @TerraDrone_JP

数週間後に日本を訪れます。イカれたハードウェア系の創業者はどこにいますか？

@gabriel1 I think it’s actually sine modulated and you’ll reach it in 2044!

we're on a linear explosion in revenue, spearheading ourselves to a predicted $100m ARR by 2045 this puts us on good track to become the worlds slowest startup to reach $100m ARR

@JensHonack Test subjects =his wife IIRC

Reading Shannon’s “Mathematical Theory of Communication”: Brother literally built an English LLM proxy by distilling prediction frequencies from his test subjects (page 7). lol wow

@magnushambleton @albertwenger I guess the simple answer is that it’s just much easier to do targeted research on machines than humans.

@albertwenger It's very frightening to me that we have created systems of this level of intelligence and with so many similarities to how our own minds work, before we have a deeply technical understanding and definition of what consciousness even is

Another exhibit for why we shouldn't simply dismiss machine consciousness

@NikoMcCarty Can recommend house of huawei! Apple in China was a revelation though

A Reading Guide to Chinese Biotech Some friends and I visited China to learn about its biotech ecosystem. Beforehand, we assembled a list of 51 "high signal" books, podcasts, and articles to help us prepare. During the trip, we also asked people who *they* followed for their news on China, and then we added many of those names to this guide. Hopefully this will be a useful resource for folks who are planning to visit (which I recommend, especially if you work in drug development.)

Office of the day! Drove 7h today, saw a precision metal stamping company (more on that soon). Now Token- and Euromaxing until the end of the day to get a demo ready for tomorrow!

@Fritz_Espenlaub True, I just have it ausgeschaltet for German

@mountbranch Kind of relying on twitters autotranslate feature

ich will bisschen mehr yimby/tpot/tech Leute in München finden! Ein paar kenne ich natürlich, aber wen gibt es noch, den ich vlt nicht auf dem Schirm haben könnte? Am liebsten direkt ganze Netzwerke, oder coole Projekte :)

Meta-analysis finds that taking an ice bath after strength training impairs gains. The inflammation, damage, stress is the SIGNAL for adaptation. Take it away...we blunt the response. Sometimes, damage is a good thing...

Sunny day: boring Sunny day with a dark purple cloud: interesting

Why is it ‘cancelled’ in the U.K. but ‘canceled’ in the U.S.? Because we gave them that L in 1776.

@magnushambleton Disagree, venture capitalist (or the Swedish riskkapitalist) means venturing into risk. The micro act of investing into one company should be ultra high risk, but on a portfolio level naturally lower risk

Although most startups fail, ”founder” as a job category is not the same, because it’s inherently uncertain and is meant to be. A founder is a risktaker taking a huge swing. But a VC is supposed to be able to manage and reduce that risk via their portfolio

VC is a weird job because the vast majority of people who are called VCs are never going to actually do their job (to return 3x+ on the capital they manage).

@Fritz_Espenlaub Ähnlich in Stockholm, alle viel zu gleichmäßig cool.

In den hippen Vierteln Berlins sieht man ausschließlich perfekt durchgestylte Leute. Jeder ist unique und lässig und die Lässigkeit ist von oben bis unten bis ins letzte Detail durchkalkuliert. Nichts ist dem Zufall überlassen. Beeindruckend, aber auch bisschen langweilig.

surprised this works. Would love to see inference times go down