Thom Kozik @thomkozik

@sama Pot?… Meet Kettle.

@lubeartiddy “My eyes are up HERE…”

Nope. Not entirely. The underlying training data, which isn't known, still carries influence. Community projects have shown that you can mitigate for a lot, but you're never going to fully "fix" that aspect of these models. The reality is that other than a few enterprises desperate to save on token costs, are going to bother with the effort/trouble of fine-tuning and de-censoring on all these models pouring out of China that are climbing the benchmark charts. And therein I believe lies the whole purpose. For decades, China’s economic strategy has been to flood markets with cheap knockoffs that the greedy Western capitalists can’t afford to say “no” to. They have done this time and time again in almost every single market sector one can name. Why should this be any different?

Chinese open-source LLMs (DeepSeek, Qwen, GLM, etc.) perform well, but Westerners are willfully oblivious to the CCP’s role & influence shaping them. Benchmarks are impressive, but so is just how naïve Westerners making use of these models are about the formal role the CCP takes in shaping their strategy and ideology. By law, these companies are *required* to have an internal CCP Committee to make sure certain that the work produced fits within party objectives and priorities, right down to political indoctrination study sessions on Chairman Xi ‘s Thoughts for the staff. The internal CCP Committee also shapes; • Personnel & leadership alignment: Manage/advise on Party member development, key appointments, and ensuring leadership stays aligned with Party priorities. • Compliance & content guardrails: Shapes how models handle “sensitive” topics per Chinese law — aligning with socialist core values, national security rules, and CCP content requirements. It affects data curation, RLHF/safety training, and what the base model refuses or biases toward. • Chinese law requires generative AI to avoid content that violates “core socialist values,” endangers China’s image, national security, or incites subversion. This applies to the company’s platforms even when accessed from outside China. The weights may be public, but the organizational incentives, reinforcement learning, and guardrails they applied aren’t.

Legacy Media types are calling this Alex Karp interview a “crash-out” so that’s your first clue that he is actually saying something extremely insightful. He is articulating what real “AI safety” looks like in the enterprise. Not abstract alignment research or certification by a government-run DMV for AI. Real AI safety for businesses is the ability to control their own data, model weights, and compute — so a frontier lab can’t hoover up their proprietary knowledge and turn it into their next product. As Karp explains, technical customers want “control over their compute, their models, their data stack, and their alpha. They want to know they own the means of production, and it’s not being transferred to someone else.” Don’t think that can happen? Just look at Figma. According to The Information, Anthropic “blindsided” its then-business partner with the launch of Claude Design. Figma’s founder said Anthropic had not been “consistently honest” with them. Anthropic’s chief product officer had even served on Figma’s board until three days before the launch of Claude Design. Figma’s stock has fallen sharply this year while Anthropic’s valuation has surged. This isn’t an isolated example. Anthropic has launched Claude Science, Claude Security, Claude Legal, and of course Claude Code — each expanding into categories previously served by companies building on top of their models. The pattern is consistent: watch where value is being created, then move in directly. Dominate the model layer, then use that position to capture the most lucrative verticals. Dario has argued that open source models powerful enough to compete with Anthropic are “dangerous.” But dangerous to whom? Not to enterprises that want to retain control over their data and workflows. Dangerous to a business model that benefits from customers having few real alternatives at the model layer. As Karp exposes, true enterprise safety isn’t trusting that a lab’s future roadmap won’t include your business. It’s retaining the ability to choose — at the model layer — who gets to see and use your alpha.

Palantir @PalantirTech

2 weeks ago

Palantir CEO Alex Karp on what customers actually want, the real business of frontier labs, and the importance of open source models: “What the technical customers want is control over their compute, their models, their data stack, and their alpha. They want to know they own the

375 752 6K 4.5M 2K

@Wolfways44 @sfliberty

If you are on the verge of AGI or ASI, why isn’t your model smart enough to recognize espionage distillation in real time? You say “cure cancer in a few years.” Isn’t sniffing illicit distillation quite a bit easier than curing cancer? Why write letters to DC? Just use AGI.

@tricalt We'd love to see that playbook and how we can apply it to our agentic emergency preparedness infrastructure

@jdoliner @altryne @midjourney

Key points missing: - Meta’s data center does not *consume* 75M annually, and holds NO Rio Grande water rights. Their usage comes from the municipality’s well system. The OP also selected the highest year of the past several, and neglects to note that up to 30% is returned to the wastewater treatment systems, which in turn gets recycled or returned to the water table. snopes.com/fact-check/zuc…

The real winner in the Anthropic fight is open source AI

Lin Qiao @lqiao

a month ago

x.com/i/article/2066…

52 105 732 1.3M 1K

@DanPMorgan @pmarca If Starmer really cared about the adverse effects of social media, why is he allowing politically acceptable social media like Blue Sky to still be unfettered?

@RoKhanna @elonmusk @JoeBiden

@PabluGablu @wholemars @RoKhanna @neeratanden @elonmusk @JoeBiden And funnily enough, after the EV tax credit was removed, Tesla still sold more cars than before.

@tonitrades_ @RyanWatkins_ $TAO

@RoKhanna @BernieSanders It's honestly astonishing that we have to waste brain electricity even thinking about retarded communists in 2026. Grow up.

The next war won't be won by armies, navies or air forces alone. It'll be won by the country whose 19 year olds can code, whose factories can build drones in weeks not years, and whose grid stays on when someone tries to switch it off. Industry. Society. Economy. That's the fight now. We're not ready. And we're not being honest about what getting ready will cost.

Milton Friedman's greatest regret. The federal government discovered the perfect crime in 1943: make employers collect taxes before workers ever see their paychecks. You think you earn $60,000 per year, but you actually earn $75,000 and hand over $15,000 to politicians without ever touching it. The psychological difference is enormous. Before payroll withholding, Americans wrote quarterly checks directly to the Treasury. Picture yourself sitting at your kitchen table, writing a $3,750 check to the IRS every three months. The pain was immediate and visceral. Politicians faced constant pressure to justify every dollar because citizens felt the extraction in real time. Withholding transforms this concrete loss into an abstract accounting entry. Your employer becomes an unpaid tax collector, and you never experience the actual cost of government. Worse, most people celebrate their tax refunds as government generosity rather than recognizing them as interest-free loans they provided to politicians. The Treasury collects your money throughout the year, spends it immediately, then returns your own cash and receives gratitude. This system enables the explosion in government spending you witness today. Defense contractors billing $640 for toilet seats, agricultural subsidies for corn syrup, and congressional salaries for 535 people who rarely show up to work. When taxation feels painless, voters stop demanding accountability for how their money gets spent. Milton Friedman helped design withholding as a wartime emergency measure and later called it his greatest regret. Free market economists recognized that the psychological pain of direct taxation creates political pressure for fiscal restraint. The temporary always becomes permanent in government hands, and the emergency justification disappears while the extraction mechanism remains forever.

Out of 16.4 million Americans who served in WWII, only about 40,000 are still alive. They’re dying at a rate of ~100 per day. These are the heroes who saved the world from tyranny. Find one. Thank one. Listen to their stories. While you still can.

@TheRealTroyWade @elonmusk The answer is right there in your comment; "We can't even keep up" Demand -- even excluding data centers -- is outpacing new supply.