Alexandra @AI__Alexandra

👇

David Sacks @DavidSacks

12 hours ago

About 8 months ago, I warned that “Anthropic is running a sophisticated regulatory capture strategy based on fear-mongering.” This take was controversial at the time; now look how many people are saying it.

237 553 7K 485K 588

THE FULL ENDING OF GAME 4. UNCUT. 🍿🍿🍿🍿

Anthropic clearly believes that words and information can be "unsafe". This is the type of thinking upstream of safetyism and cancel culture, classifying snippets of code as weapons, jailing people for free speech or writing open source code. The exact opposite type of thinking produced the Web, Wikipedia, Linux, Bitcoin and Ethereum, and ironically, much of the historical AI research that led to this moment. I have faith in the tendency of the universe to produce the most entertaining outcome, so I cannot believe this is the real AI takeoff it's just too lame.

Por allá por Oaxaca existe un plato llamado Tlayuda con una tortilla de maíz gigante, casi igual que las tortillas de harina sobaqueras

Brown Panther @drilliebobby

15 hours ago

Nobody in the tortilla industry has the balls to make a corn tortilla the size of a flour tortilla

231 1K 34K 1.1M 679

@espn Insane

Scientific research is fundamental to advancing civilization and helping people globally to solve the most critical problems, from medicine to materials, from brain science to physics, and much beyond. This is only possible when scientists have access to the best tools of the time to conduct scientific research, including having access to AI-based tools.

"What concerns me about Amodei and Anthropic in particular is the consistent pattern of being singularly focused on being the one winner with all of the power, with limited consideration of how everyone else may react to that situation." This is a thoughtful piece on all this.

Stratechery @stratechery

3 months ago

Anthropic and Alignment Anthropic is in a standoff with the Department of War; while the company's concerns are legitimate, it position is intolerable and misaligned with reality. stratechery.com/2026/anthropic…

105 92 689 445K 394

One of my favourite art supply shops in Paris: Magasin Sennelier. It supplied materials to Picasso and Edgar Degas. A must-see for art lovers.

keyboard_arrow_left Previous keyboard_arrow_right Next

@StevenLevy Perhaps your thoughts on the most pressing issue in the AI community: x.com/eliebakouch/st…

elie @eliebakouch

2 days ago

mythos will be bad ON PURPOSE on ai "frontier llm research" tasks, this is very very sad for the research community also the fact that this is un purpose not visible to the user is crazy

355 636 6K 2.6M 1K

As believers of open research, we are disappointed to see Anthropic silently degrading Fable 5 for AI development "Any topic related to building pretraining pipelines, distributed training infrastructure, or ML accelerator design... may have limited effectiveness through Claude via methods such as prompt modification, steering vectors, or parameter-efficient fine-tuning." Not only do they get to decide what you use LLMs for in research, but this also enables them to silently intervene in your research without you knowing. This sets a dangerous precedent. If a model refuses openly, users can understand the boundary. If a model falls back to another model, users can still evaluate the difference. But if a model silently modifies or weakens its own answers while still pretending to help, researchers lose the ability to know whether a failed result came from their own idea, their implementation, or an invisible intervention by the model provider. That is not safety. Safety policies should be transparent, auditable, and user-visible. On top of that, the people most harmed by this are not the largest labs with massive teams and proprietary infrastructure. It is the independent researchers, academic groups, startups, and open-source builders who rely on public tools to compete, innovate, and pioneer AI for everyone else.

Great summary of what my feed looks like right now, worth a read if yours is not covering all this already.

Robert Scoble @Scobleizer

a day ago

"Misanthropic." I've never seen the AI community so angry at a major new model release. I asked my AI (an agent that @blevlabs made for me) to gather all the backlash. +++++++++++ THE BACKLASH AGAINST CLAUDE FABLE 5'S RESTRICTIONS The best analysis of why this matters:

112 204 1K 216K 408

The AI safety discourse may finally be entering its “wait, are we making the models worse?” era.

SITUATION DETECTED: Google DeepMind has released DiffusionGemma, an open model that generates entire blocks of text simultaneously rather than word-by-word, achieving up to 4x faster output on dedicated GPUs.

@Spotify Eclectic

@Velissse Fit and material quality.

@Wendys At what point do you decide to sell your chili sauce? This is urgent. Thanks.

@EvaLovesDesign Love these

Why does a jet of air suddenly bend and cling to a curved surface? That’s the Coandă Effect. A moving fluid creates a pressure gradient that allows the flow to follow a curved surface instead of continuing in a straight line. As it bends, it also pulls in surrounding fluid, a process known as entrainment. This relationship shows the pressure gradient required to keep a fluid moving along a curved path. From aircraft aerodynamics to industrial fluid systems, the Coandă Effect is one of the most fascinating examples of how pressure shapes motion.

Gaussian Process Deep Learning combines the uncertainty quantification of Gaussian processes with the representation-learning power of deep neural networks. A Gaussian process (GP) defines a distribution over functions, allowing predictions to be accompanied by principled uncertainty estimates. Unlike standard neural networks, which typically produce point predictions, GP-based models naturally capture confidence and uncertainty in their outputs. The key idea is to learn hierarchical representations while retaining the Bayesian nonparametric nature of Gaussian processes. Deep Gaussian Processes stack multiple GP layers, enabling highly flexible function approximation and richer representations than classical GPs. In statistics, Gaussian process deep learning provides powerful tools for nonparametric regression, classification, uncertainty quantification, and Bayesian inference. In machine learning, it is used for active learning, Bayesian optimization, surrogate modeling, and learning from limited data. In deep learning, GP methods improve calibration, robustness, and interpretability while providing theoretical connections between infinitely wide neural networks and kernel methods. In reinforcement learning, Gaussian process models are valuable for model-based RL, exploration under uncertainty, and sample-efficient learning. Since uncertainty estimates guide exploration, agents can make better decisions with fewer interactions. The broader insight is that intelligence requires not only prediction but also awareness of uncertainty. Gaussian process deep learning offers a principled framework that combines expressive representation learning with rigorous probabilistic reasoning. Image: share.google/fdmkvG2M78osQL…

I asked Fable 5 to recreate Monopoly but make each of the properties an AI lab or startup. It implemented everything - game rules, money system, turns, even share codes for multiplayer games. Once you have a monopoly, you can build racks and eventually a data center 🤓