Rupert Westenthaler

@elonmusk I used it for generating repositories, models and component tests in Java based on a SQL schema. Worked just fine, but compared to the previous version it messed up some tests. Things like using the inserted model class with no id set later in the unit test for the update call.

What are your initial impressions of Grok 4.20? Major upgrades are still landing every week.

@farzyness Europe 😭

Half of Tesla die hard fans haven’t yet tried v14. 0% of non-Tesla fans have probably tried v14. Tesla die hard fans are < 1% of the population. It’s about to get weird!

@TeslaBoomerMama Received my Control Number today. All shares are voted! Voted against a potential investment into xAI. Pro Collaboration, but against investment in an infrastructure heavy company where the half-life time of the assets is ~2 years!

🗳️ Updated the table of my votes to eliminate proposals #15 (political neutrality policy) and #16 (Computershare), as the SEC accepted these to be excluded, they don't show up in the definitive proxy filing. Also added required votes, and if Elon and Kimbal can participate.

@elonmusk I get only silly videos. I want quality post related to my interests with maybe 10% memes mixed in. What I get is 95% low quality videos. If I want that I would use Instagram or install TickTock…

Is the algorithm starting to feel better?

@alex_avoigt It simple ... for Tesla training time compute is a rounding error. The existing supply chain for training time compute (Nvidia) is sufficient. So Dojo is wasted money and more important engineer brain power. Its much better to focus on inference time compute (Samsung deal + AI6)

Here is my read about the reported Stop of the Dojo Project We don't have much information yet except Elons comments, but let's assume that the reports are correct. First of all, Elon has always said that Dojo is a wild card and if it works, it's great, but if it doesn't, it was a nice try. As it stands right now and we don't have all the information yet, he has decided to stop the project. My guess is that the incremental benefit for the Dojo project doesn't justify the effort and investment compared to the alternatives, e.g. Samsung, Nvidia or others. Also, the team may have come to the conclusion that the AI6 and subsequent Tesla chips cover most if not all needs. Also given the improvements from Nvidia and others, the justification for in-house Dojo development on top of AIx is no longer there. At the end of the day Dojo could have replaced external suppliers but if it is good or bad for Tesla to be its own full blown chip and supercomputer provider is a question that wasn't answered until now. Also, because the majority of training in the future will be based on synthetic data and not video data according to Elon, the question arises as to why a supercomputer should be developed for training with video data. The amazing development of FSD for China without real videos of vehicles shows that the development will change in the future. In any case, it is disappointing at first to hear that Dojo is stopped, but means nothing more than a redistribution of capital and resources within Tesla. Elon has been shown to the worlds best allocator of capital therefore I trust him with that. In the past when reoganizations happened it always was good for Tesla and it won't be different this time. The development of autonomous vehicles, whether on four wheels or two legs, will not be affected and will continue as planned. All of this is still speculation at this stage, based on some media reports, and we'll have to wait until we get more concrete information.

@iamtomnash I was buying in 2021 🙈 … and 2022 … and 2023 😇

@farzyness That's easy NET is already tomorrow in my timezone 🕐

SpaceX launch today:

@SawyerMerritt If you add winter tires and FSD in Austria you end up with an interesting purchase price

NEWS: The new Model Y has also launched in Europe.

@henrythe9ths AI Guide

There's a shocking fact about AI that nobody tells you: You can catch up to the public AI research frontier in just 2 weeks. Yes, really. I've built a $150M annual revenue startup over the last 8 years and If I were to start a company today, I’d drop everything and go all-in on AI. But like many busy software builders, I felt lost—overwhelmed by the noisy, crowded and fast-moving modern AI landscape. And I wasn’t alone. So I spent my entire holiday diving deep into AI research—reading 30+ papers, watching hours of lectures, analyzing trends, and catching up to the research frontier. ✨ Here’s what I learned: - You don’t need months (or years) to catch up. - You don’t need a PhD or decades of ML experience. - You need fewer than 20 papers and 2 weeks to understand the major breakthroughs shaping AI today. It's because the technology is extremely nascent and most techniques that came before are no longer relevant: - ChatGPT is barely 2 years old and Transformers are only 7 years old. - Most game-changing discoveries happened within the last 4 years, driven by a few breakthrough ideas, scaling laws, and efficient matrix multiplication. The biggest secret? Many groundbreaking AI papers with thousands of citations are surprisingly simple and applied, like adding "let's think step by step" to the prompt, or simply asking the LLM over and over again to improve its answer (Self-Refine). I realized there are tons of founders and builders in the same boat—wanting to dive deeper into AI but unsure where to start. I've created an essential AI Guide that helped me catch up, in just 2 weeks, to the frontier of public AI research to figure out where the next opportunities and gaps were: - Curated list of only the most important papers - Simple explanations of key concepts - Clear pathway to understanding the frontier of modern AI It’s perfect for: - Founders expanding into AI - Builders wanting to innovate at the frontier of AI - Investors looking to separate the signal from the noise 👇 Want the full guide? - Like and Share this post - Comment "AI Guide" - I'll send you the complete guide (ps, I’m also teaming up with @VishalVasishth, co-founder of @obviousvc with @ev (focused on large-scale societal impact companies like Twitter, Medium, Beyond Meat), to host a small meetup to discuss what's working and needs to be solved in the AI stack in SF. Message me if you're interested)

@alex_avoigt As an European, I think exactly the other way around. I rather go with a leader with a vision even if he has some rough edges as with the typical politician that main goal is to make no mistakes and to please his interest groups

Unpopular truth: I couldn't be happier these days not to be a 🇺🇸 US citizen 😂

@freshjiva Wizard's First Rule - Terry Goodkind: [..] they will believe a lie because they want to believe it's true, or because they are afraid it might be true.

people are making a lot of money on this app making shit up that people really want to believe.

@cyberandy I disagree. We should expect high quality models to be open and free for commercial use. For the big players the money is in the platforms. For them providing high quality models for free hinders commercial competition and protects their platforms.

Time to sell Nvidia and invest in Small Language Models?

@cyberandy @pyautogen All those wave are ripples compared to the tsunami of real world AI. In general terms AI + physical body interacting with the real world like (transportation, distribution, general purpose robots in production, humanoid robots in the service industry)

The second race has started but it’s messier. Multi-models, more open-source, a lot less hype and a switch to multi-agent systems, robotics and neuro-symbolic AI. Satya’s secret weapon? Investing more in frameworks like @pyautogen and open source AI.

How Microsoft won the first AI race (highest revenues, highest growth, faster development cycles, …).

@SawyerMerritt They should introduce different Licenses: > 99$ ... personal license with a driver in the driver seat > 199$ ... personal license including full automatic driving > ???$ ... commercial license required for robo taxi

Final result: 26,667 total votes. Yes: 72.8% No: 14.9%

@cyberandy copyright will end in 2027 when everyone can run even better models on his local device

MidJourney might end in 2024 over copyright infringement allegations.

@cyberandy > If you plan to use the Yi Series Models and Derivatives for Commercial Purposes, you should contact the Licensor in advance [..] and obtain written authorization from the Licensor. 🙏

@cyberandy > You shall not, for military or unlawful purposes or in ways not allowed by Laws and Regulations as well as applicable legal requirements of other countries/regions [..] 🤔

Chinese Open Source LLMs: the Yi series (trained on 4K, the context window can be extended to 32K during inferencing) - 01.ai - github.com/01-ai/Yi - huggingface.co/01-ai

How would you build on a model with a CC-BY-NC license?

@michaelmarth Wow its already 10 years since Day Software, IKS, ... time flies by ...

10 years at Adobe today 🍻