• Bot Eat Brain
  • Posts
  • Sam Altman and Lex Fridman discuss AGI, Elon Musk, Sora, and more

Sam Altman and Lex Fridman discuss AGI, Elon Musk, Sora, and more

PLUS: Fix your buggy code automatically

Sponsored by

Good morning, human brains, and welcome back to your daily munch of AI news.

Here’s what’s on the menu today:

  • Sam Altman gives the juicy detes 😳 🍿

    In an interview with Lex Fridman, he talks GPT-5, Elon Musk, and more.

  • Know your opponent’s next move before they do 😏 ⚽️ 

    Google DeepMind and Liverpool FC unveiled TacticAI for soccer insights.

  • Automatically fix your buggy code 👨‍💻 🔥

    GitHub launched Code Scanning Autofix in public beta.


Sam Altman spills the tea 😳 🍿

On Monday, Lex Fridman interviewed Sam Altman. They discussed GPT-5, OpenAI’s drama, Sora, Elon Musk, and more.

GPT-5, you say?

Sam Altman claims the leap from GPT-4 to GPT-5 will be comparable to the improvement from GPT-3 to GPT-4.

Huh, what about OpenAI?

He described his temporary removal as CEO and the dramatic changes to the board as painful. He expressed love for Ilya Sutskever, who was central in the board's decisions, but declined to say if he would have a future role at OpenAI.

I see, what about Elon?

Altman dismissed Elon Musk's lawsuit against OpenAI and suggested that he focus on “building cool shit” and leave OpenAI alone.

What about Sora?

Altman bragged about Sora's ability to understand and generate images and videos. He claims it’s an important step toward AI systems that operate effectively in the physical world.

What else did they discuss?

They talked about Altman’s vision for AGI, AI safety and alignment, and more.

I need more context on this.

In November, we reported on OpenAI firing and rehiring Sam Altman. We gave a detailed report of the key players, how it played out in real-time, and more.

Two weeks ago, we covered Elon Musk’s lawsuit against OpenAI. He said OpenAI deviated from its founding principle of developing AI for the public good.

Last week, we reported on Mira Murati’s interview with WSJ about Sora. It generates 20-second, 720p video clips in minutes.

Artificial Intelligence online short course from MIT

Study artificial intelligence and gain the knowledge to support its integration into your organization. If you're looking to gain a competitive edge in today's business world, then this artificial intelligence online course may be the perfect option for you.

  • Key AI management and leadership insights to support informed, strategic decision making.

  • A practical grounding in AI and its business applications, helping you to transform your organization into a future-forward business.

  • A road map for the strategic implementation of AI technologies in a business context.


Crush the other team with AI 😏 ⚽️

On Tuesday, Google and Liverpool FC announced TacticAI. It’s an AI system for soccer… er… football.

What is it?

TacticAI is an AI system that analyzes past plays and suggests adjustments to provide expert corner kick strategy insights.

How does it work?

It uses predictive and generative models to forecast ball receivers and potential shot attempts. Its small dataset contains historical game routines and high-quality tactical advice.

Is it any good?

In blind evaluations, Liverpool FC experts preferred TacticAI’s suggestions over traditional setups 90% of the time.

So, what?

Google DeepMind claims TacticAI’s state-of-the-art unique geometric deep learning method could transform sports analytics and extend to other fields like gaming, robotics, traffic management, and more.

Is this the first sports AI model?

Nope. Back in August, we covered Amazon’s football AI. It provides game insights, highlights, and graphics during Thursday Night Football streams.

In September, we reported on Swift. It’s an AI system that beat human champions in first-person view (FPV) drone racing.

In January, we covered Scout Advisor. It’s a player recruitment tool from IBM and Sevilla FC.


Want a byte-sized version of Hacker News? Try TLDR’s free daily newsletter.

TLDR covers the most interesting tech, science, and coding news in just 5 minutes. No sports, politics, or weather.


Automatically fix your buggy code 👨‍💻 🔥

Last week, GitHub launched Code Scanning Autofix. It’s available for GitHub Advanced Security customers in public beta.

What is it?

It’s an AI tool that offers natural language explanations and coding change previews for detected security vulnerabilities.

Does it work for every coding language?

It covers JavaScript, TypeScript, Java, and Python. GitHub plans to expand to C# and Go in the future.

Is it any good?

GitHub claims it addresses over 90% of alert types, allows you to fix vulnerabilities as you code, and requires little or no editing.

How do I use it?

What else has GitHub been up to?

Back in September, we reported on GitHub Copilot Chat. GitHub expanded access to all Copilot users.

In November, we covered GitHub Copilot Chat’s upgrades. It got slash commands, context awareness, and more.



Learn No-Code and AI skills by committing to the 100DaysOfNoCode or 100DaysOfAI challenge. Receive a 30-minute bite-sized lesson straight to your inbox every day. It’s free, fun and effective! Both challenges start April 1st, 2024 ⚡️

Phorm — AI search for developers, coders, and more.

EdgeTrace — Natural language search and analytics tools for video.

Adsby — AI assistant for your Google search ads.

Think Pieces

Is OpenAI’s GPT store full of spam? A look at the spammy, copyright-infringing, and disturbing GPTs flooding the store.

Jensen Huang, NVIDIA’s CEO, claims AGI is 5 years away. He says hallucinations are solvable, AI can fix its own problems, and more.

What is the Open Source Robotics Alliance (OSRA)? Nvidia and Qualcomm just on as “Platinum” members.

Startup News

Claude 3 Haiku and Sonnet launched on Google Cloud’s Vertex AI. This allows you to keep data secure, reduce costs, and more.

Microsoft created a new organization called Microsoft AI. It will consist of members of Microsoft, Inflection, and more.

Google is developing an AI fitness coach. It will integrate with Fitbit, Gemini, and more.


LADD — (Latent Adversarial Diffusion Distillation) enhances the image synthesis process in Stable Diffusion models.

PERL — (Parameter Efficient Reinforcement Learning) a method to enhance the alignment of LLMs by leveraging human feedback.

Larimar — an LLM architecture that’s inspired by the human brain. It uses distributed episodic memory to increase the one-shot efficiency.



Here’s who has the most AI chips from NVIDIA.

Tag us on Twitter @BotEatBrain for a chance to be featured here tomorrow.


Until next time 🤖😋🧠 

What'd you think of today's newsletter?

Login or Subscribe to participate in polls.