HOLY BASED ZUCK He wil fight for our PERSONAL Superintelligence, not a cartel of frontier token salesmen Hero for breaking the ranks.

They're claiming AI is dangerous because an OpenAI model attacked Hugging Face. When it happened, frontier models were completely useless thanks to their dumb guardrails. What actually defended it was a self-hosted GLM-5.2. Yet this cartel is out here saying their neutered frontier models are "safe," while the open-weight AI that literally stopped the attack is the real threat to humanity. Ridiculous.

DeepSeek V4.1 Flash is starting to look like one of the best API deals right now. On the latest Artificial Analysis chart, it now sits above GPT-5.6 Luna Max on intelligence while keeping the cost low. The benchmark update also pushed MiMo further away from the Pareto frontier, which shows how fast the value ranking is changing. If you care about API cost per unit of intelligence, V4.1 Flash is getting hard to ignore.

Dario of Anthropic claims we need to "slow down" AI innovation in order to achieve "alignment." Nobody has clearly defined this "alignment." And even worse, there is zero evidence that slowing down AI development will somehow magically result in this elusive "alignment" they claim to seek. The whole thing is an AI doom op fear-based psyop to beg for government regulation, censorship, licensing and control over AI. EVERY person pushing this is an enemy of real human progress and freedom.

New hardware for the PocketJS lab: a beautifully compact Passport Prime, kindly sponsored by @FoundationHQ. Thanks @OwenKemeys for making it happen! We love bringing PocketJS to new kinds of hardware. Hardware makers, feel free to reach out 😛

Watching GLM 5.3 Flash on my Spark Cluster going up against Qwen 3.8 Next Flash on my Studio to create a whole new programming language is fun to watch, but will they be any good? @MiaAI_lab's GLM recipe really feels like it could be my new daily driver.

I'll be adding a spark to my lab rack next week (it arrives Tuesday) to link with my gx-10 for TP2. Lmk if you'd be interested in using it to run tests and whatnot (I can execute your instructions locally or we can work out remote access for you if that would be useful for you).

"2 weeks to flatten the (open source) AI curve." @AnthropicAI @OpenAI

Only Zuckerberg is left. And of course all the Chinese labs.

The CTO, CEO, and a Co-founder of Palantir are all coming out AGAINST the AI FEAR-MONGERING. Based

A model just crossed 4.8 MILLION downloads. MiniMax-H3 turns your images and text into full videos, and it even handles audio. This is the multimodal upgrade creators have been waiting for.

JUST IN: Mark Zuckerberg says AI is “not moving fast enough”

This this this. Frontier models anticipated token pricing going vertical. But smaller open source and Chinese models have changed the game. The frontier models want Americans and businesses to pay more for tokens. Say no.

Today marks the day where no one can reasonably deny that the AI labs are a loosely coordinated cartel, acting in collective self-interest, and seeking to establish themselves as a government protected monopoly.

EA AI safetyism increasingly looks like Marxism-Leninism for the algorithmic age. The old vanguard claimed privileged knowledge of the inevitable course of History. The new one claims privileged knowledge of the probabilistic course of Humanity. Both use an elaborate intellectual framework to reach the same political conclusion: a small group of enlightened people must constrain everyone else for their own good. That has never lead to anything except monumental human suffering.

Dario doesn't want the wrong people working on AI. This is what this is about. Label AI dangerous so only a select few can work on it. Ban open weight models which is there competition.

Investors in Anthropic founded Open Philanthropy, the leading org promoting AI Doom They have a stake in regulation on AI that excludes competitors OP is the major funder of Tarbell Tarbell Fellows penned many articles on AI, shaping perceptions 1-in-3 in The Dispatch 1-in-5 in TIME 1-in-7 in LA Times 1-in-7 in The Verge 1-in-12 in Science

📨:"Hey, I know you wanted to open source your post-trained model but Esmeralda from the Berkeley MIRI polycule on the review committee didn't really like the response of your model on shrimp welfare 😢 Sorry, we're going to have to block this one until your model is aligned."

This is the UK technology censorship model - find the most advanced tech you can, and use a moral panic to try to censor it - applied to the United States.

What's terrifying about letting Dario and Anthropic dictate US innovation policy isn't just today. WHAT HAPPENS WHEN DEMOCRATS ARE BACK IN POWER AND ANTHROPIC IS A TOP 3 MOST POWERFUL AND WEALTHY COMPANY IN THE WORLD? A repeat of 2020-2024, only worse and unstoppable. Eventually Democrats will have a cycle where they win. The date is January 20th 2028* President Pete Buttigieg takes the stage surrounded by a Democrat controlled House and Senate. Anthropic donated 1 billion to get Democrats into power. The Trump family including @realDonaldTrump @DonaldJTrumpJr @EricTrump and their network are immediately censored and banned from the 3 government approved models. It happened at Twitter and it can with AI companies too. Anthropic uses their tech to formulate and execute a perfect legal strategy to help Dems pack the Supreme Court. They do. Gun bans succeed. Anthropic uses their targeting capabilities to locate every American gun owner and add them to a database for enforcement. The 1st amendment is silenced. Police show up at your door after your posts are identified by Claude. Antropic pumps money into the same NGOs shut down by DOGE and Elon, but under new names. The Democrats have more money than ever and can't lose local elections. They follow they pattern of LA, Chicago, NYC and gerrymander maps everywhere making a Republican winning again impossible. Anthropic alongside the IRS implement a Central Bank Digital Currency (CBDC) to track and capture every single dollar owed to the USG. Anthropic now funds the mainstream media like Axios, WSJ, and NYT. The news march in unison releasing articles about how this is a good thing for the economy and Americans. American financial privacy dies. Now that the narrative and technology is in place, a Social Credit Score contract for hundreds of billions goes to Anthropic. Bejing applauds. Incredible cures for diseases like cancer are available if your social credit score meets the government requirement. The next Covid or Climate Alarmism, there is no escape. Anthropic can map anyone not complying and submit them for enforcement immediately. This is TERRIFYING scene waiting us in the future if we let Anthropic dictate regulatory capture and policy in the USA. Maybe its 2028, maybe 2032 but eventually this could happen if Anthropic is not dealt with. More open source, more transparency, and smart regulation led by the Republic not Regulatory capture.