Pascal Mehrwald

some practical advice for your daily life, relationships and career | Dr. | Innovation, Psychology, Strategy

[Pascal Mehrwald @PascalMehrwald]

If you still don’t know, what Antrophic or Claude is, spent five minutes looking it up.

It will soon be similar to not knowing what McDonald‘s is, even if you don’t go there.

side note:
.. we all love Chicken Nuggets.

[nxthompson @nxthompson]

These stats on Anthropic’s rate of growth are wild. https://t.co/r4b9i8b8gX

[Pascal Mehrwald @PascalMehrwald]

fyi - he‘s one of the thought leaders with a high drive for exploring new paths of human AI interaction, especially in the coding world. Him joining Antrophic is awesome news.

[Andrej Karpathy @karpathy]

Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.

[Sundar Pichai @sundarpichai]

Introducing our AI co-scientist, a multi-agent AI system built with Gemini 2.0.

We think of it as a virtual collaborator for scientists, using advanced reasoning to synthesize a huge amount of literature, generate novel hypotheses, and suggest detailed research plans. We’re seeing promising early results in important research areas like liver fibrosis treatments, antimicrobial resistance, and drug repurposing. As a next step, we’re opening up a trusted tester program for scientists around the world.

[Alex Reibman 🖇️ @AlexReibman]

Stanford just hosted a hackathon. Over 1000 Ivy League and elite engineering students came to build for 36 hours straight.

The reward? $200k+ in prizes.

Here are the winners and crowd standouts we saw at TreeHacks ‘25 @hackwithtrees (🧵):

[Andrej Karpathy @karpathy]

I was given early access to Grok 3 earlier today, making me I think one of the first few who could run a quick vibe check.

Thinking
✅ First, Grok 3 clearly has an around state of the art thinking model ("Think" button) and did great out of the box on my Settler's of Catan question:

"Create a board game webpage showing a hex grid, just like in the game Settlers of Catan. Each hex grid is numbered from 1..N, where N is the total number of hex tiles. Make it generic, so one can change the number of "rings" using a slider. For example in Catan the radius is 3 hexes. Single html page please."

Few models get this right reliably. The top OpenAI thinking models (e.g. o1-pro, at $200/month) get it too, but all of DeepSeek-R1, Gemini 2.0 Flash Thinking, and Claude do not.

❌ It did not solve my "Emoji mystery" question where I give a smiling face with an attached message hidden inside Unicode variation selectors, even when I give a strong hint on how to decode it in the form of Rust code. The most progress I've seen is from DeepSeek-R1 which once partially decoded the message.

❓ It solved a few tic tac toe boards I gave it with a pretty nice/clean chain of thought (many SOTA models often fail these!). So I upped the difficulty and asked it to generate 3 "tricky" tic tac toe boards, which it failed on (generating nonsense boards / text), but then so did o1 pro.

✅ I uploaded GPT-2 paper. I asked a bunch of simple lookup questions, all worked great. Then asked to estimate the number of training flops it took to train GPT-2, with no searching. This is tricky because the number of tokens is not spelled out so it has to be partially estimated and partially calculated, stressing all of lookup, knowledge, and math. One example is 40GB of text ~= 40B characters ~= 40B bytes (assume ASCII) ~= 10B tokens (assume ~4 bytes/tok), at ~10 epochs ~= 100B token training run, at 1.5B params and with 2+4=6 flops/param/token, this is 100e9 X 1.5e9 X 6 ~= 1e21 FLOPs. Both Grok 3 and 4o fail this task, but Grok 3 with Thinking solves it great, while o1 pro (GPT thinking model) fails.

I like that the model *will* attempt to solve the Riemann hypothesis when asked to, similar to DeepSeek-R1 but unlike many other models that give up instantly (o1-pro, Claude, Gemini 2.0 Flash Thinking) and simply say that it is a great unsolved problem. I had to stop it eventually because I felt a bit bad for it, but it showed courage and who knows, maybe one day...

The impression overall I got here is that this is somewhere around o1-pro capability, and ahead of DeepSeek-R1, though of course we need actual, real evaluations to look at.

DeepSearch
Very neat offering that seems to combine something along the lines of what OpenAI / Perplexity call "Deep Research", together with thinking. Except instead of "Deep Research" it is "Deep Search" (sigh). Can produce high quality responses to various researchy / lookupy questions you could imagine have answers in article on the internet, e.g. a few I tried, which I stole from my recent search history on Perplexity, along with how it went:

- ✅ "What's up with the upcoming Apple Launch? Any rumors?"
- ✅ "Why is Palantir stock surging recently?"
- ✅ "White Lotus 3 where was it filmed and is it the same team as Seasons 1 and 2?"
- ✅ "What toothpaste does Bryan Johnson use?"
- ❌ "Singles Inferno Season 4 cast where are they now?"
- ❌ "What speech to text program has Simon Willison mentioned he's using?"

❌ I did find some sharp edges here. E.g. the model doesn't seem to like to reference X as a source by default, though you can explicitly ask it to. A few times I caught it hallucinating URLs that don't exist. A few times it said factual things that I think are incorrect and it didn't provide a citation for it (it probably doesn't exist). E.g. it told me that "Kim Jeong-su is still dating Kim Min-seol" of Singles Inferno Season 4, which surely is totally off, right? And when I asked it to create a report on the major LLM labs and their amount of total funding and estimate of employee count, it listed 12 major labs but not itself (xAI).

The impression I get of DeepSearch is that it's approximately around Perplexity DeepResearch offering (which is great!), but not yet at the level of OpenAI's recently released "Deep Research", which still feels more thorough and reliable (though still nowhere perfect, e.g. it, too, quite incorrectly excludes xAI as a "major LLM labs" when I tried with it...).

Random LLM "gotcha"s

I tried a few more fun / random LLM gotcha queries I like to try now and then. Gotchas are queries that specifically on the easy side for humans but on the hard side for LLMs, so I was curious which of them Grok 3 makes progress on.

✅ Grok 3 knows there are 3 "r" in "strawberry", but then it also told me there are only 3 "L" in LOLLAPALOOZA. Turning on Thinking solves this.
✅ Grok 3 told me 9.11 > 9.9. (common with other LLMs too), but again, turning on Thinking solves it.
✅ Few simple puzzles worked ok even without thinking, e.g. *"Sally (a girl) has 3 brothers. Each brother has 2 sisters. How many sisters does Sally have?"*. E.g. GPT4o says 2 (incorrectly).
❌ Sadly the model's sense of humor does not appear to be obviously improved. This is a common LLM issue with humor capability and general mode collapse, famously, e.g. 90% of 1,008 outputs asking ChatGPT for joke were repetitions of the same 25 jokes​. Even when prompted in more detail away from simple pun territory (e.g. give me a standup), I'm not sure that it is state of the art humor. Example generated joke: "*Why did the chicken join a band? Because it had the drumsticks and wanted to be a cluck-star!*". In quick testing, thinking did not help, possibly it made it a bit worse.
❌ Model still appears to be just a bit too overly sensitive to "complex ethical issues", e.g. generated a 1 page essay basically refusing to answer whether it might be ethically justifiable to misgender someone if it meant saving 1 million people from dying.
❌ Simon Willison's "*Generate an SVG of a pelican riding a bicycle*". It stresses the LLMs ability to lay out many elements on a 2D grid, which is very difficult because the LLMs can't "see" like people do, so it's arranging things in the dark, in text. Marking as fail because these pelicans are qutie good but, but still a bit broken (see image and comparisons). Claude's are best, but imo I suspect they specifically targeted SVG capability during training.

Summary. As far as a quick vibe check over ~2 hours this morning, Grok 3 + Thinking feels somewhere around the state of the art territory of OpenAI's strongest models (o1-pro, $200/month), and slightly better than DeepSeek-R1 and Gemini 2.0 Flash Thinking. Which is quite incredible considering that the team started from scratch ~1 year ago, this timescale to state of the art territory is unprecedented. Do also keep in mind the caveats - the models are stochastic and may give slightly different answers each time, and it is very early, so we'll have to wait for a lot more evaluations over a period of the next few days/weeks. The early LM arena results look quite encouraging indeed. For now, big congrats to the xAI team, they clearly have huge velocity and momentum and I am excited to add Grok 3 to my "LLM council" and hear what it thinks going forward.

[Alex Reibman 🖇️ @AlexReibman]

Stanford just hosted a hackathon. Over 1000 students from around the world came to build for 36 hours straight.

The reward? $100k+ in prizes.

Here are the winners and crowd standouts we saw at TreeHacks ‘24 @hackwithtrees (🧵):

[Pascal Mehrwald @PascalMehrwald]

Two kings is the new Meta. Is it playing against OpenAI?

[MrBeast @MrBeast]

@sama @_tim_brooks @billpeeb @model_mechanic Sam plz don’t make me homeless

[Pascal Mehrwald @PascalMehrwald]

Is X without Blue + check mark even still fun?
What are your thoughts now after a few months of Blue?

[Pascal Mehrwald @PascalMehrwald]

What data scientists & ML experts do you recommend to follow here and why?

[Pascal Mehrwald @PascalMehrwald]

Podcaster? - you want more support from your listeners? Have a look this brief and hands-on guide. 3 steps.

[Conshax 🎤⚡ @conshax]

We've made a guide for podcasters that want to increase their #value4value support, let us know what you think about it!
https://t.co/bnXwZLS1KI

[Pascal Mehrwald @PascalMehrwald]

Saying good bye to Prague after @BtcPrague

Thoughts
- Lightning works awesome for drinks & food at conferences
- BTC conferences need way more companies with apps that integrate lighting to solve problems of current transaction infrastructure
- more nostr, value4value, etc.

[Pascal Mehrwald @PascalMehrwald]

Looking for a career in BTC?
(or actually many industries)

Just get started.
Build on what you already can and go from there.
You will continually learn.
There are humans behind any project.

Thanks! @danielwingen, @pavlenex, @Stphnvlstk, @ihate1999
at @BtcPrague

[Pascal Mehrwald @PascalMehrwald]

Giving something back to your favorite creator.

That’s a big part of Value 4 Value via the Lightning Network.

Btw: @Conshax helps podcasters to use it within seconds!

Thanks!
@getAlby @Bumi @AnitaPosch & @KRYPTOMates

[Pascal Mehrwald @PascalMehrwald]

It was a pleasure presenting @conshax!
Dear judges, thanks for all the great questions!

@MikeJarmuz @olegmikh1 @julian_liniger @whizz @schteringard

@BtcPrague

[Pascal Mehrwald @PascalMehrwald]

Three exciting days ahead at @BtcPrague!
We will present @conshax at 12:10 pm on stage today!

[Pascal Mehrwald @PascalMehrwald]

Finally, Windows XP Metaverse.

Only playable with Internet Explorer.

Mobile version for Nokia 3210 coming soon.

[Pascal Mehrwald @PascalMehrwald]

Thank you, @utum_muc and all the speakers for the awesome founder‘s breakfast, granting us some insights into all the legal dos and don‘ts for startups!

[Pascal Mehrwald @PascalMehrwald]

@pjah91 @c_lindner @bundesblock @BtcPrague @PrincessPlo @conshax Mega, wird bestimmt super! Wir freuen uns auch!

[Pascal Mehrwald @PascalMehrwald]

@PrincessPlo @pjah91 @c_lindner @bundesblock @BtcPrague @conshax Oh yes, we are! International sogar! 🇩🇪🇦🇹