Heisenberg Professor for Machine Learning. Tech enthusiast. Livin' the dream. Views are my own.

Joined May 2020
Replying to @eric_is_weird
I agree with this and I have made a lot of mistakes with it but found some solutions that work for me. 1. The first mistake was having too large a lab, where grant writing exceeded >35 hours a week, so I downsized the lab by 50% (this was not entirely voluntary - details are in this post: [1]) 2. I saw some colleagues who seemed to pick projects extremely carefully and have a good sense of what projects would be valuable. They seemed to be enjoying their job and studying things they were excited about. One prof in particular had only 2 students, but the prof was studying and creating amazing work all day and was extremely animated. 3. I realised that you can actually have some intellectually challenging "projects" that you are not paid to do, and not evaluated on, and you can do them for relaxation. Doing this reduces my level of burnout and helps tremendously. I am working on several projects that I am not publishing on, in new areas for me, that I am quite obsessed with (!). Now this is odd as it can be tempting to turn them into grants or lab projects but I have learned the hard way that it is better not to. I would encourage people to find challenging projects like these (it does not have to be science obviously) but it will help with everything else. [1] xitter.cf/PTenigma/status/199285…
I have made some major changes given the extremely chaotic NIH funding situation in the last year. 1. I downsized the lab by 35% in the last 12 months. We had a 2.5-2.8M USD/year payroll 2 years ago, which included 900K/y in faculty payroll across all grants. In the US, PhD faculty in the med school are usually paid up to 80-90% from grants, and in the recent past several had zero start-up, making it harder for them to stay solvent, so we'd put 80-90% of the asst profs' salary spread across several grants. We also have to pay 300K USD a year to rent the space our lab occupies, and I like it but we have to bring in the rent and utilities as a direct cost. These are very hard constraints as if we run out of money in any one year, we are evicted and the lab closes. So we made some hard decisions. 2. When I say downsized, several of our postdocs who were/are extremely brilliant, productive, creative people got faculty jobs and instead of finding a new postdoc I just did not hire a new person. We have an incredible consortium manager, who many of you know. And I have only one graduate student, she is an exceptional mathematician and computer scientist. So the lab was able to downsize by about 1/3 of its size per 2 years by literally not 'replacing' people and still able to do a lot of the work with several people covering and multitasking on a large number of projects. 3. We would often submit a grant with 60-90K USD/year for my lab, with me as contact PI, and 10-12 subawards to support a very complex consortium structure (think of federated deep learning, where you would have to do this). This meant we had been submitting 14 NIH R01s (or R01-equivalents) per year with a success rate of 15-20% after 19 months and 2-5 submissions. With the new rule that a PI can submit only 6 NIH grants per calendar year, we cannot submit 14 R01s per year which means I cannot submit a grant with only 90k/year for my lab as we would go bankrupt given the expected payout of 15% or less after 19 months. Conversely, we are working quite a bit harder before submitting a grant. 4. We had grants in the last year scored at 2nd percentile, 8th percentile, 13th percentile, 21 impact score, all not funded. I am not sure I want to re-submit a grant that scored 2nd percentile and was not funded. 5. We are trying to develop more technically innovative methods, which involves quite a lot more reading and studying. This has not been bad. 6. Some grants involved an insane amount of reporting, I mean just literally ridiculous with meetings all the time that did not seem to be of any use to anyone. I ended my participation in 3 of the grants and contracts that had the most meetings, and things got massively better as it freed up ~10 hours a week. I think people should be mindful of how many meetings they have, because it can be the last straw for many collaborations, unless people have a lot of time on their hands and enjoy the social aspect of it. 7. I have studied how very famous PIs and new and successful PIs run their labs, and even just kind people who are doing well and seem to be fair to everyone. For one of them, it was shocking how little grant writing they do, but I think it was made easier as the faculty salaries are paid by the university in Europe and they seemed to pick projects very carefully and have a good sense of what projects would be valuable. They even seemed to be enjoying their job and studying things they were excited about.
6
6
34
4,442
That!
Science is not some hobby that serves the self actualization of people in a publicly sponsored job placement program. It is the systematic search for truth.
1
136
TimHahn retweeted
Science is not some hobby that serves the self actualization of people in a publicly sponsored job placement program. It is the systematic search for truth.
Wow, over 1,000 mathematicians have now joined the Association for Human Mathematics (AHM), committing to: > NOT publish AI generated math papers > NOT work with frontier labs > 654 of them went even further to NOT even use AI in their research, at all
142
332
46
3,362
93,328
I am sincerely confused. How are those statements arguments? "Intelligence is not consciousness" How is that applying to machines more than humans? "Biology is essential" and "The physical organization may matter more than the computation." That's the question!
Spectacular Nature piece on why AI is not conscious and probably will never be. Three arguments: 1) Intelligence is not consciousness. Intelligence is about solving problems. Consciousness is about having subjective experiences. AlphaFold can solve super difficult problems. By contrast, our brains performs complex operations entirely outside awareness. Problem-solving and consciousness are two separate things. 2. Biology is essential. Our brains are alive, down to every cell. Their activity is inseparable from the living tissue that produces it. The idea that consciousness can be reproduced simply by running the right computations on silicon components who do not fight for their life is a very far fetching and, I would add, almost religious, assumption. 3) The physical organization may matter more than the computation. According to integrated information theory, consciousness requires specific causal relationships between a system’s physical components. The computer architectures running today’s AI lack this organization. The comparison with AlphaFold is particularly revealing: Almost nobody things AlphaFold is conscious, despite its inner working is very similar to an LLM Why? Because it does not speak. LLMs do speak like us and therefore we fall into an illusion: Since they speak like us, they must be like us. But this inference is wrong. LLMs are just a series of matrix multiplications. They are not conscious, and they will never be. * Paper in the first reply
27
3
3
91
7,775
Zeit für Basics: Preise steigen bei Knappheit und locken neues Angebot an.
Hohe Preise zeigen, wo etwas fehlt.
Feste Preise machen leere Regale.
Wer Schlange steht, zahlt mit Zeit.
Was nichts kostet, wird verschwendet.
Niemand weiß besser als der Käufer, was er braucht.
Das Wissen steckt in vielen Köpfen, nicht in einem.
Tausend Versuche finden mehr als ein kluger Kopf.
Tausch macht beide reicher.
Konkurrenz macht billiger.
Wo man wählen kann, muss man gut sein.
Ein Laden ohne Kunden schließt.
Fehler kosten den, der sie macht.
Wer nicht verlieren kann lernt nichts.
Wer selbst zahlt, schaut genauer hin.
Fremdes Geld gibt sich leichter aus.
Was niemandem gehört, pflegt niemand.
Wer für Leistung nichts bekommt, leistet bald weniger.
Gute Absichten ersetzen keine Preise.
Um aus der Wirtschaftskrise herauszukommen, braucht Deutschland einen radikalen Kurswechsel – hin zur geplanten Marktwirtschaft. Das empfiehlt der Ökonom Tom Krebs in seinem neuen Buch „Kapital und Krise“, das er zusammen mit Surplus im Basecamp in Berlin vorgestellt hat.
25
51
5
270
6,211
Sometimes, as a scientist, you have to wonder how a well-argued, but completely fact-free opinion 1) makes it into such a prestigious journal and 2) is then treated like evidence by much of the media.
Spectacular Nature piece on why AI is not conscious and probably will never be. Three arguments: 1) Intelligence is not consciousness. Intelligence is about solving problems. Consciousness is about having subjective experiences. AlphaFold can solve super difficult problems. By contrast, our brains performs complex operations entirely outside awareness. Problem-solving and consciousness are two separate things. 2. Biology is essential. Our brains are alive, down to every cell. Their activity is inseparable from the living tissue that produces it. The idea that consciousness can be reproduced simply by running the right computations on silicon components who do not fight for their life is a very far fetching and, I would add, almost religious, assumption. 3) The physical organization may matter more than the computation. According to integrated information theory, consciousness requires specific causal relationships between a system’s physical components. The computer architectures running today’s AI lack this organization. The comparison with AlphaFold is particularly revealing: Almost nobody things AlphaFold is conscious, despite its inner working is very similar to an LLM Why? Because it does not speak. LLMs do speak like us and therefore we fall into an illusion: Since they speak like us, they must be like us. But this inference is wrong. LLMs are just a series of matrix multiplications. They are not conscious, and they will never be. * Paper in the first reply
49
I’m sorry but those arguments are nonsense absent an explanation of what consciousness is. It’s like saying planes don’t fly because they don’t flap their wings.
Spectacular Nature piece on why AI is not conscious and probably will never be. Three arguments: 1) Intelligence is not consciousness. Intelligence is about solving problems. Consciousness is about having subjective experiences. AlphaFold can solve super difficult problems. By contrast, our brains performs complex operations entirely outside awareness. Problem-solving and consciousness are two separate things. 2. Biology is essential. Our brains are alive, down to every cell. Their activity is inseparable from the living tissue that produces it. The idea that consciousness can be reproduced simply by running the right computations on silicon components who do not fight for their life is a very far fetching and, I would add, almost religious, assumption. 3) The physical organization may matter more than the computation. According to integrated information theory, consciousness requires specific causal relationships between a system’s physical components. The computer architectures running today’s AI lack this organization. The comparison with AlphaFold is particularly revealing: Almost nobody things AlphaFold is conscious, despite its inner working is very similar to an LLM Why? Because it does not speak. LLMs do speak like us and therefore we fall into an illusion: Since they speak like us, they must be like us. But this inference is wrong. LLMs are just a series of matrix multiplications. They are not conscious, and they will never be. * Paper in the first reply
50
13
198
7,313
"You reduce crime by eliminating poverty." This is one of the greatest intellectual confusions of our century. It sounds compassionate. It feels obvious. And the moment you actually test it, it falls apart. Let's look at the data. First, the correlation is real. Nobody serious denies that poorer areas have more crime. In Sweden, kids from the poorest 20% of families are about 7 times more likely to be convicted of a violent crime than kids from the richest 20%. But correlation is not causation. So Swedish researchers did something brilliant: they compared siblings. Same parents, same home, same upbringing, but different family income at different points in their childhood, because parents' income moves over time. Result, on more than 500,000 children (Sariaslan et al., British Journal of Psychiatry, 2014): once you account for what's shared inside the family, the effect of income on violent crime goes from 7x to nothing. The hazard ratio drops to 0.95. No effect. Finland ran the same test on 650,000 people. Same result: the income-crime link disappears when you compare siblings. What predicts crime isn't the bank balance. It's what travels inside families. Second test: what happens when poverty spikes for everyone at once? 2008. US unemployment doubled from 5% to nearly 10%. 15% of Americans officially poor. If poverty creates crime, crime should have exploded. Instead it fell. Year after year. Robbery, the most "economic" crime of all, dropped 10% in 2010 alone. Criminologists literally said they were stumped. Same story in the Great Depression: 25% unemployment, and crime went down in many cities. Third test: take families out of poverty and see. The US government ran a randomized lottery called Moving to Opportunity. Thousands of families in public housing got vouchers to move to low-poverty neighborhoods. For boys, violent arrests dropped at first, property crime arrests went up, and over time the effects faded. New zip code. Same behavior. Fourth: the claim that police don't matter. This one is just false. Chalfin et al. (American Economic Review: Insights, 2022): each additional police officer prevents about 0.1 homicides. And per capita, the lives saved are twice as many among Black victims as among white victims. The people who suffer most from "police don't matter" are the poorest neighborhoods. So what does predict crime? The Dunedin study followed 1,000 children from birth to age 32. Self-control measured between ages 3 and 11 predicted adult criminal convictions, even after controlling for social class and IQ (Moffitt et al., PNAS, 2011). Cognitive ability matters too: across 700,000 Swedish men, lower cognitive scores at 18 track with violent convictions, and the link barely moves when you control for childhood socioeconomic status (Frisell et al., 2012). And here's the part nobody quotes: in Dunedin, kids whose self-control improved over time did better than predicted. Behavior can change. Which means responsibility is real. Why does this matter? Because "poverty creates crime" isn't just wrong. It's toxic. It turns criminals into victims. The man who robbed you becomes a symptom of "the system," and you, the real victim, become a footnote in someone else's narrative. It merges two completely different things. "You're poor, we'll help you" is one policy. "You're poor and you committed a crime, so we'll excuse you" is another. Fusing them insults the millions of poor people who work, raise their kids, and never hurt anyone. The vast majority of poor people are not criminals. That alone kills the theory. It removes responsibility. Build a political system on "your circumstances made you do it" and you destroy the one thing that actually changes behavior: the certainty that you own your choices. Every excuse the state manufactures is an incentive. And it feeds the state forever. If every crime is a social failure, every crime justifies a new program, a new budget, a new agency. Crime becomes the business model of bureaucracy. The alternative is simple and old. A lighter state that protects instead of explains. The same rules for everyone, from the projects or from a gated community. Certain, swift punishment for anyone who commits a crime. And never, ever telling someone their crime was written in their social condition. Treating someone as responsible is respecting them. Excusing someone because they're poor is the most condescending thing you can do to them.
This quoted post is unavailable.
215
1,793
183
8,845
251,204
TimHahn retweeted
So many people are going to get one-shotted by super persuasive super intelligences and I might be one of them
6,334
5,184
1,387
82,078
8,251,235
That!
Replying to @troywhiteyyc
2026-31: AI agents master research/coding; peptide/gene therapies mainstream for aging/disease. Fusion prototypes scale commercially. Crewed Mars landings succeed. 2031-36: AGI arrives, economies shift to abundance models. Clean energy dominates. Permanent lunar bases online. 2036-41: Neural interfaces widespread. Asteroid mining starts. Healthy lifespans near 100. 2041-46: Multiplanetary expansion. Human-AI hybrids common. Tech-driven climate stabilization advances.
56
TimHahn retweeted
Replying to @troywhiteyyc
2026-31: AI agents master research/coding; peptide/gene therapies mainstream for aging/disease. Fusion prototypes scale commercially. Crewed Mars landings succeed. 2031-36: AGI arrives, economies shift to abundance models. Clean energy dominates. Permanent lunar bases online. 2036-41: Neural interfaces widespread. Asteroid mining starts. Healthy lifespans near 100. 2041-46: Multiplanetary expansion. Human-AI hybrids common. Tech-driven climate stabilization advances.
95
82
20
1,081
310,588
I thought it'd be a good idea to do a quick summary of the points I have made about the problems in the foundations of physics, just to make it more difficult for the offended physicists on X to pretend misunderstanding me: (A) The problem with the foundations of physics is that physicists continue to use methods that demonstrably do not work. They have made tens of thousands of so-called "predictions" for new particles, forces, extra dimensions etc etc, that were simply wrong. You can check this yourself. The literature is full of this nonsense. It has come to the point where a "prediction" in this field no longer means anything. It's embarrassing. And they're not doing anything about it. I have NOT said that the problem is that it takes them a long time to make progress. It is true of course, that that theory development in the field has basically stalled since the completion of the standard model. But who is to say whether this is just because the questions are difficult. If you think I have said that it is taking too long, you are confusing me with someone else. I have said instead that the problem is that they are using methods that evidently do not work. And that they do not update these methods is hugely worrisome because it demonstrates a failure of science to self-correct. It shows that the self-correction in this community is broken. (B) I have been very clear with saying that the problem is theory development in the foundations of physics. Not the foundations of physics per se (clearly experiment has made progress), not physics per se (clearly there is more to physics than the foundations), but theory development in the foundations of physics. (C) In my book Lost in Math, I explain that historically the method that has worked best (for theory development in the foundations) has been requiring mathematical consistency. This is why I argue physicists in these areas should prioritize solving problems of mathematical consistency and/or inconsistency with experiments. Examples: quantum gravity, measurement problem in quantum physics, dark matter. This is not an exhaustive list. (D) I did not say that mathematical consistency is either sufficient or necessary for a successful new theory. I made a much weaker point, which is that historically it has worked well. Evidence speaks for it. Obviously just because a theory is mathematically consistent doesn't mean it also correctly describes nature. Meaning: Physicists must also make sure their theory can, and will be tested. If they work on a theory without thinking about how to experimentally test it, that is not scientific. (E) People who haven't read my book sometimes think I say physicists use too much maths. This could not be farther from the truth. I argue they need to take maths more seriously. (F) Because of the above, I have repeatedly said that string theory has a good motivation. If you think I am a big critic of string theory, again, you are confusing me with someone else. (G) I have said that while these issues seem specific to the foundations of physics, they are symptoms of a bigger underlying problem that affects all of academia, which is that scientists get rewarded for producing papers that they themselves call progress. Progress in academia has no tangible external measure, hence they are optimizing irrelevant goals. It's a systemic problem that has been much discussed by sociologists and economists and many, if not most, academics are to some extent aware of it. In the foundations in particular, physicists have hit on the rather obvious simple fix of producing theories that simply have no relation to the real world and then discussing them forever, claiming that paper production counts as progress in some sense. I'm not having it and neither should you. This nonsense needs to stop. (H) I have been writing and speaking about this for about 20 years. Indeed it's how I got to know @ericweinstein , because I organized a conference about this a long time ago. The idea that either of us is going on about this because it is somehow "popular" contradicts facts. Rather what has happened is that at last at least the more informed public is paying attention to what we have been pointing out for a long time is a huge problem.
169
215
38
1,672
71,346
TimHahn retweeted
Dit is nicht jut sach ich mal...
Seit Anfang 2022 hat der industrielle Bereich rund 600.000 Beschäftigte verloren, während im staatsnahen Bereich fast 700.000 hinzukamen.
17
10
1
245
4,228
TimHahn retweeted
Honestly: Photoshop and other creative apps are childs play. The fun starts when the big guns get into focus: Highly technical CAD applications with license costs in the 5, 6 or 7 digits per seat per year. Cadence Xcelium, Innovus and Virtuoso. Synopsys DesignCompiler, PrimeTime, Synplify. Siemens Questa. Altium. AutoCAD. I recently built an open source Schematic Tracer that is objectively better than anything that Cadence or Synopsys have ever done. In 3 days. They're all cooked.
AI has made desktop apps commercially non-viable. This industry died when this free, open source photoshop clone was announced 5 days ago. 7000 people have already started working on it. 1500 improvements in 5 days. All free. It was a good run. github.com/storytold/photocr…
214
288
31
4,488
159,088
Terance Tao: Imagine prescribing a drug no one understood. Medical community:
113
891
82
13,026
292,710
TimHahn retweeted
Replying to @sparbuchfeinde
Ja, die Differenz stimmt weitgehend: In DE liegen Luft-Wasser-Wärmepumpen inkl. Einbau oft bei 23–40.000 €, in FR bei 12–20.000 €. Gründe sind höhere Lohnkosten und längere Montagezeiten, 19 % MwSt (vs. 5,5 % in FR), strengere Normen (Fundamente, Elektroanschluss, Abgleich), oft komplexere Systeme und die prozentuale Förderung, die Preise stabilisiert.
4
13
2
243
15,173
really started questioning if I want to move back to Germany at all it would great for family (my son being close to grandparents again etc) but everything else just keeps on telling me i'm not really welcome as an entrepreneur the latest berlin election is another sign that people just haven't learned anything from the past we need innovation and growth but all signs point in the other direction it feels like the economy needs to get so bad that people literary feel poor again but this could take some time it makes me very sad, this is my home country, my native language and i still feel deeply connected to it I want my kids to have grandparents around but it also makes me so angry to see a country just butcher it's economy I want to feel valued in a society by contributing as an entrepreneur it's entrepreneurs who create jobs, who create change, who drive countries foward but it really is a mind virus that most Germans have lost their appreciation for this and only want the Nanny State to become bigger it will just lead to more entrepreneurs and companies leaving to more pro business countries global mobility has never been easier it's all so tiresome really
129
46
14
941
47,154
Jeder Beschäftigte in Deutschland muss das vollständige Arbeitgeberbrutto monatlich überwiesen bekommen und selbstständig alle Steuern und Abgaben überweisen. Wir würden von heute auf morgen mental in einem anderen Land leben. In einem Land, in dem selbst dem Mindestlöhner auffallen würde, dass er über 40 Prozent seines Einkommens an den Staat abdrücken muss. In einem Land, in dem ich nicht mehr an Tischen mit acht politisch überdurchschnittlich gebildeten Akademikern sitzen muss, die mich beim Versuch, das Konzept des Arbeitgeberbruttos zu erklären, fassungslos anschauen. nius.de/kommentar/warum-nehm…
94
642
14
4,138
53,908
Linken-Chefin Ines Schwerdtner bei Maischberger: "Da würde zumindest mal einiges fürs Gemeinwohl zusammenkommen, wenn wir Elon Musk ordentlich besteuern würden!" Elon Musk soll endlich etwas zum Gemeinwohl beitragen? Seine Unternehmen beschäftigen bereits rund 158.000 Menschen direkt. Sie zahlen Milliarden an Löhnen und Steuern, bauen Autos, Raketen, Satellitennetze, KI-Infrastruktur und Medizintechnik. Hinzu kommen vermutlich Hunderttausende indirekte Arbeitsplätze bei Zulieferern. Vielleicht sollten wir aufhören, Gemeinwohl ausschließlich daran zu messen, wie viel Vermögen der Staat jemandem wegnehmen kann.
152
580
21
4,267
53,493
Replying to @patrickc
i quote a statement from a famous founder in Europe which I met 2 days ago, (Herman Hauser / ARM-Processor) 👉 “Out of 10 start-ups I supported and helped scale in the past, NOT A SINGLE ONE has remained in Europe. Today, they are all multibillion-dollar companies that are now contributing to strong GDP growth outside the EU — predominantly in the United States." nothing to add.
1
21
829