# Factual errors are more dangerous than weak visual judgment

Markdown copy of https://realalpha.app/s/factual-errors-are-more-dangerous-than-weak-visual-judgment/ on Real Alpha, the best minutes from long interviews with the people building AI. Cite and link the page, not this copy.

- Speaker: [Joseph Kim](https://realalpha.app/people/joseph-kim/) · Joseph Kim is a leads ChatGPT for Financial Services at OpenAI.
- Host: [Shawn Wang (swyx)](https://realalpha.app/people/swyx/) · Co-founder, AI Engineer · Wrote “The Rise of the AI Engineer” and founded Smol AI, maker of AI News.
- Show: Latent Space
- Episode: [Building ChatGPT for Finance — Joseph Kim, OpenAI's Head of Financial Services](https://realalpha.app/v/v-d1WTqIULs/), Oct 31, 2026
- Moment: 12:32 to 15:49 of the episode (3 min), playing on https://realalpha.app/s/factual-errors-are-more-dangerous-than-weak-visual-judgment/; the episode from there: https://realalpha.app/v/v-d1WTqIULs/?t=752

## Quote

> “the career-ending mistake is something where you get a piece of information factually incorrect. And that piece of factual information **changes the conclusion**.”
> — [Joseph Kim](https://realalpha.app/people/joseph-kim/) at [13:59](https://realalpha.app/v/v-d1WTqIULs/?t=839) · Latent Space

## Takeaways

1. A factual error matters most when it changes a conclusion.
2. Review layers reduce the chance errors reach senior decision-makers.
3. Presentation taste remains a model weakness.

## About this segment

Joseph Kim distinguishes career-threatening factual errors from weaker judgment about presentation quality. He says model reliability on facts is improving, while taste still needs work.

From “Building ChatGPT for Finance — Joseph Kim, OpenAI's Head of Financial Services”, Oct 31, 2026. [Watch the full episode](https://realalpha.app/v/v-d1WTqIULs/)

## Transcript

Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there.

[12:36](https://realalpha.app/v/v-d1WTqIULs/?t=756) I'd say there's this like sort of misconception where in the finance world, one mistake really compounds like your whole career is at risk if you submit something wrong. How should people think about getting started and using it? So if you're an analyst, how should you use these tools? And do you have like any analogies towards where we're at right now? So there's like levels of autonomy and self-driving.

[12:57](https://realalpha.app/v/v-d1WTqIULs/?t=777) One of your slides, I think said like 2027, you see no more PowerPoint. A lot of people will be using AI tools in the loop. So how should people think about where we are now and where things are going and how should they adopt these tools? If they have that misconception of like, one mistake, my career is over. Obviously there's human in the loop, but how should they also think about it going forward as they need to offload more and more?

[13:18](https://realalpha.app/v/v-d1WTqIULs/?t=798) Yeah, no, that's wonderful. I have two perspectives on it and they start with different assumptions. I think the notion that my career is over, if you make one mistake, is an incorrect notion. And I have a line of thought that follows that. But let's first assume that it's true. And in the case that it's true, what I would say is not all mistakes are created equally, and not all model capabilities are created equally.

[13:40](https://realalpha.app/v/v-d1WTqIULs/?t=820) And so when you think about the mistake that a model or a human can make, that could be a career ending, it's usually once you get up to a certain level. If you're a VP, the stakes are higher than when you're a junior analyst, because there's a presumption of work having been checked up to that point. So if an MD sees a mistake, that's a huge problem. A lot of the safety nets have failed for that mistake to make it up to an MD, right?

[14:01](https://realalpha.app/v/v-d1WTqIULs/?t=841) But what I would say is the career-ending mistake is something where you get a piece of information factually incorrect. And that piece of factual information changes the conclusion. That is obviously the biggest type of mistake that you can make. And what I would say is we're seeing a lot of progress in attention to detail. and the hallucination used to be a big term that used to come up in 2022, 2023 that doesn't come up nearly as often nowadays.

[14:25](https://realalpha.app/v/v-d1WTqIULs/?t=865) I would say that we're starting to see huge improvements in the reliability of models as it pertains to factual mistakes. So the kind of egregious mistake that would get you fired, models are getting much better at and continue to get much better. And so I feel actually quite confident that a combination of that continual improvement plus the existing system checks in place will make it such that that type of mistake, the chance of that propagating up to an MD is just very low probability, right?

[14:51](https://realalpha.app/v/v-d1WTqIULs/?t=891) What I would say the models are lacking today and they still need to improve is actually not the career-ending mistake, but it's the bit that involves human judgment and taste. So it's the, this slide looks ugly, right? And it's not wrong, but it's formatted, like the colors look a little bit off, something about it looks sloppy, and therefore, you know, this is not a good slide. and that's not a career ending mistake in the way that a factual mistake is but it still matters it still matters in terms of how useful that work is and so as we think about the progression going into 2027 what I would say is I expect the model to get much better at the human taste component and that's what I was kind of alluding to in my keynote is I think the model will get much better at human taste on things that are softer objective

[15:37](https://realalpha.app/v/v-d1WTqIULs/?t=937) functions, it's not is it right or wrong it's comparative, is there something that's better than this. And I think models are going to get much better along that dimension moving forward. How do we solve taste? How do we train taste into the models?