Claims Browser

Even criminals with very limited skills will be able to target victims at every scale due to AI

"Even criminals with very limited skills will be able to target victims at every scale"

Anthropic will seek a public valuation of $2 trillion or more

"The Financial Times reported Anthropic will seek a public valuation of $2 trillion or more."

Grok 4.7 will be ready in the next 3-4 weeks

"Elon says Grok 4.7 will be ready in the next 3-4 weeks"

Engineers will spend more time managing the machines that do the work

"what happens when engineers spend more time managing the machines that do the work"

Transformative AI could reshape power, governance, the economy, and individual freedom

"transformative AI could reshape power, governance, the economy, and individual freedom"

OpenAI will continue work as the EU AI Act advances

"The work will continue as the EU AI Act advances"

The future belongs to specialized models built on private company data rather than general-purpose models

"the future belongs to specialized models built on private company data rather than general-purpose ones"

Big tech companies might be more screwed than expected given how toxic data centers have become

"at this point the big tech companies might be more screwed than I thought, given how toxic data centers have become"

Models will keep absorbing the harness into their weights until it becomes a harness for human attention rather than for the model

"Models keep absorbing the harness into their weights — soon, it will be a harness for human attention rather than for the model."

Anthropic's revenue in Q2 may not recur at the same scale next year after the collapse of tokenmaxxing, as companies like ATT turn to open source models to save costs

"THE big issue for Anthropic is that the kind of revenue they were making in the second quarter of this year may well not recur, at the same scale, next year, now that, in the aftermath of the collapse of tokenmaxxing, companies like ATT are increasingly turning to open source models from other vendors, in order to save costs."

If OpenAI goes down, the damage could spread widely across the AI industry

"Still, as Ed Zitron and I have both long noted, if OpenAI goes down the damage could spread"

Future RL training will incorporate current-generation AI systems to provide reward signals for next-generation AI

"This is our current best-guess model of future AI development, and so the role of debate is to ensure that the LLM judgements used during training provide as accurate a reward signal as possible."

LLM judgments will become an increasingly important aspect of frontier RL training, especially for long agentic trajectories

"The trend of increasing test time compute will likely only exacerbate this problem. Even for tasks like coding, there are many aspects of desirable LLM agent behavior that are fuzzy, and so LLM judgements are likely to become an increasingly important aspect of frontier RL training."

Advancements in autonomous systems and machine intelligence may soon enable states to project force and collect revenue without relying on human cooperation, potentially undermining the fundamental bargain between power and people that has sustained individual freedom throughout history

"Advancements in real-world autonomous systems may soon mean that states will be able to project force—domestically and internationally—without relying on cooperative soldiers or police officers. And advancements in machine intelligence may mean that the state can collect the revenue it needs not from the fruit of individual human labor, but from the outputs of data centers."

Power may no longer require a society-wide bargain to be sustained, and the people may have only a small seat at the negotiating table or none at all

"Power may therefore no longer require a society-wide bargain to be sustained, and the people, broadly conceived, may have only a small seat at the negotiating table, or perhaps none at all."

The compute needed to be at the frontier of current model recipes is going vertical, and will only become more evident as the world accelerates along recursive self improvement

"The compute needed to be at the frontier of the current model recipe is going vertical, and as the world accelerates along the axis of Recursive Self Improvement this will only become more evident."

Human level capabilities of intelligence will be fully commoditized by open source models, while super intelligence will likely not be

"We believe human level capabilities of intelligence will be fully commoditized by open source models, while super intelligence will likely not be."

Open models will likely fill a long-tail ecosystem focused on efficiency and specialization rather than competing with closed models on the most valuable areas

"I put this mentally as my most likely outcome – open models are still incredibly useful, but fill a long-tail ecosystem relative to the closed counterparts that have monopoly ownership stakes in the most valuable areas like knowledge work collaboration, drug discovery, SWE, etc."

The established pretraining, midtraining, post-training lexicon may shift to pretraining, reasoning training, and post-training

"This could go so far as to change the established pretraining, midtraining, post-training lexicon that has been standard for a few years. It could come to be something closer to pretraining, reasoning training, and post-training."