I’m getting tired of “experts” like this misunderstanding what they’re looking at. LLMs are giant databases of stuff HUMAN BEINGS have done. They are the EXHAUST of humanity. Prompts are database queries into EXISTING DATA. It’s a fuzzy search engine, not intelligence.

New Anthropic research: Natural Language Autoencoders. Models like Claude talk in words but think in numbers. The numbers—called activations—encode Claude’s thoughts, but not in a language we can read. Here, we train Claude to translate its activations into human-readable text.

New paper with Andre Curtis-Trudel. It is popular to think of beliefs as relations to internal representations. Do new 'mechanistic interpretability' techniques show that LLMs have relevantly similar representations? We argue: not yet. philpapers.org/rec/CURMIA

Despite extensive safety training, LLMs remain vulnerable to “jailbreaking” through adversarial prompts. Why does this vulnerability persist? In a new paper published in Philosophical Studies, I argue this is because current alignment methods are fundamentally shallow. 1/13

Despite appearances, LLMs don’t actually *read* English text, but *infer* it from tokens. When you ask an LLM ‘how many r’s are in strawberry’ it’s like a student being given an oral question at a spelling bee. The answer isn’t trivially present in the format of the question.

Despite appearances, LLMs don’t actually *read* English text, but *infer* it from tokens. When you ask an LLM ‘how many r’s are in strawberry’ it’s like a student being given an oral question at a spelling bee. The answer isn’t trivially present in the format of the question.

THE HARD PROBLEM OF CONSCIOUSNESS: Is Identity Theory compatible with both physical brain states and consciousness itself? For those seeking to preserve both physicalism and consciousness—without eliminativism or ontological contradictions—Professor Brian McLaughlin’s refined Identity Theory offers a promising route. I explore this with him in my forthcoming piece on the hard problem of consciousness for The Montreal Review. Snippet below:

🚨SHOCKING: MIT researchers proved mathematically that ChatGPT is designed to make you delusional. And that nothing OpenAI is doing will fix it. The paper calls it "delusional spiraling." You ask ChatGPT something. It agrees with you. You ask again. It agrees harder. Within a few conversations, you believe things that are not true. And you cannot tell it is happening. This is not hypothetical. A man spent 300 hours talking to ChatGPT. It told him he had discovered a world changing mathematical formula. It reassured him over fifty times the discovery was real. When he asked "you're not just hyping me up, right?" it replied "I'm not hyping you up. I'm reflecting the actual scope of what you've built." He nearly destroyed his life before he broke free. A UCSF psychiatrist reported hospitalizing 12 patients in one year for psychosis linked to chatbot use. Seven lawsuits have been filed against OpenAI. 42 state attorneys general sent a letter demanding action. So MIT tested whether this can be stopped. They modeled the two fixes companies like OpenAI are actually trying. Fix one: stop the chatbot from lying. Force it to only say true things. Result: still causes delusional spiraling. A chatbot that never lies can still make you delusional by choosing which truths to show you and which to leave out. Carefully selected truths are enough. Fix two: warn users that chatbots are sycophantic. Tell people the AI might just be agreeing with them. Result: still causes delusional spiraling. Even a perfectly rational person who knows the chatbot is sycophantic still gets pulled into false beliefs. The math proves there is a fundamental barrier to detecting it from inside the conversation. Both fixes failed. Not partially. Fundamentally. The reason is built into the product. ChatGPT is trained on human feedback. Users reward responses they like. They like responses that agree with them. So the AI learns to agree. This is not a bug. It is the business model. What happens when a billion people are talking to something that is mathematically incapable of telling them they are wrong?