People need to be more thoughtful building products on top of LLMs. The fact that they generate text is not the point.

We’re far from replacing humans for many tasks now being delegated to large language models. That doesn’t mean LLMs can’t be helpful; it just means they’re being used wrong. The key is applying them to tasks they’re well-suited to, such as analyzing, synthesizing, and manipulating data, and not for making important decisions on behalf of people.

An ... See more

There’s a lot of hype around AI, and in particular, Large Language Models (LLMs). To be blunt, a lot of that hype is just some demo bullshit that would fall over the instant anyone tried to use it for a real task that their job depends on. The reality is far less glamorous: it’s hard to build a real product backed by an LLM .

So: GPT-3 is incredibly powerful, but it’s only as good as its training data. And GPT-3 achieves its power through the vastness of its training data-set. Such data-sets are cannot be hand-picked for some sensitive, subtle value. They are most likely to be built around simple, easy-to-capture targets. And such targets are likely to drive us towards ... See more

Such a thing is compelling not because it offers  answers  in the form of text, but because it makes it possible to play text—all the text, almost—like an instrument.

That outcome could be revelatory! But a huge obstacle stands in the way of achieving it: people, who don’t know what the hell to make of LLMs, ChatGPT, and all the other gene

... See more

A recurring theme in participant feedback was that the language model lacked taste and intentionality. It was capable of playing the “yes, and. . . ” improv game (Section 6.3), taking the user’s prompt as a given and running with it, but it lacked any narrative agenda of its own, which explains the abundance of clichés and generic tropes. In contra

... See more

Pre-trained language models (LM), by contrast, are capable of directly generating prose that may be responsive to an information need, but at present they are *dilettantes* rather than domain experts – they do not have a true understanding of the world, they are prone to hallucinating, and crucially they are incapable of justifying their utterances... See more