LLMs are bullshitters. But that doesn't mean they're not useful.

Note: This is a personal essay by Matt Ranger, Kagi's head of ML In 1986, Harry Frankfurt wrote On Bullshit. He differentiates a lying from bullshitting: Lying means you have a concept of what is true,...

What's Wrong with Bullshit

Past philosophical analyses of bullshit have generally presented bullshit as a formidable threat to truth. However, most of these analyses also reduce bullshit to a mere symptom of a greater evil (e.g. indifference towards truth). In this paper, I introduce a new account of bullshit which, I argue, is more suited to understand the threat posed by bullshit. I begin by introducing a few examples of “truth-tracking bullshit”, before arguing that these examples cannot be accommodated by past, process-based accounts of bullshit. I then introduce my new, output-based account of bullshit, according to which a claim is bullshit when it is presented as or appears as interesting at first sight but is revealed not to be that interesting under closer scrutiny. I present several arguments in favor of this account, then argue that it is more promising than past accounts when it comes to explaining how bullshit spreads and why it is a serious threat to truth.

LLMs believe false statements even after explicit warnings that they're false

Fine-tuning tests show "bias... toward confidently representing the claims as true."

Against False Indulgences in LLM Alignment

Unnecessary moral concern, and the false indulgences that drive it.

How Do We Talk about LLMs?

By Jim Clifford It is never fun to watch friends argue. Generative AI has created multiple fractures across our universities and the community of historians. The stakes are high and we are all deal…

Peter Gostev on Twitter / X

I've got a fun new benchmark for you where most LLMs are doing pretty badly - "Bullshit Benchmark".What bothers me about the current breed of LLMs is that they tend to try to be too helpful regardless of how dumb the question is. So I've built 55 'bullshit' questions that don't… pic.twitter.com/4o4quN5EFR— Peter Gostev (@petergostev) February 24, 2026

There's Something Fundamentally Wrong With LLMs

LLMs aren't trained on the "vast majority of speech," experts warn, a major blind spot that could have sweeping consequences.

On Bullshit

Over one million copies sold worldwideThe international and #1 New York Times bestsellerThe anniversary edition of the acclaimed book that reveals why bullshit is more dangerous than lying

Illusions of Understanding from Outsourcing Thinking to LLMs

Some illusions of understanding are an inevitable part of the research process, while others can be avoided or overcome by careful critical thinking and observation. We are facing an increased risk of avoidable illusions as more research activities are delegated to large language models (LMM). LLMs can be useful but they cannot think, and their use can undermine our thinking and understanding. Thinking for ourselves is hard and error prone but worthwhile - and there are no shortcuts to understanding.

It’s remarkably easy to inject new medical misinformation into LLMs

Changing just 0.001% of inputs to misinformation makes the AI less accurate.

Arguments in Favor of AI Fair Use

I made some notes on the nature of the training of LLMs, and about whether we as a society should consider the material used to train them as a fair use of that material.

What Google Translate Can Tell Us About Vibecoding | off the grid

There has been rather a lot of doomsaying (and perhaps astroturfing) lately about LLMs as the end of computer programming. Much of the discussion has been lacking nuance, so I’d like to add mine. I see claims from one side that “I used $LLM_SERVICE_PROVIDER to make a small throwaway tool, so all programmers will be unemployed in $ARBITRARY_TIME_WINDOW”, and from the other side flat-out rejections of the idea that this type of tool can have any utility.1 I think it best sheds light on these claims to examine them in the context of another field that’s been ahead of the curve on this: translation.

Have we been measuring AI political bias wrong? A better approach is possible.

Why ideological preferences and epistemic failure in LLMs are not the same thing — and why the difference matters