Post

... It's a text generator? Like that's its primary purpose, generating plausibly-human text from a prompt. It's both designed to do that and advertised for that. If it's not primarily a text generator then what is it for you?

I think "generative AI" is a bit of a misleading name - I think of it more as "transformative AI", the most interesting use-cases are when you pipe a bunch of text INTO the model and then instruct it to transform or process that text in some way

8/5/25, 4:11 pm1 repost8 likes

simonwillison.netMaybe Meta’s Llama claims to be open source because of the EU AI actI encountered a theory a while ago that one of the reasons Meta insist on using the term “open source” for their Llama models despite the Llama license not actually …

The simplest example is answering questions against documents - e.g. when I used it to help me find specific parts of the EU AI act relating to the term "open source" simonwillison.net/2025/Apr/19/...

Feeding in larger chunks of code and using it to help analyze potential sources of bugs is another way to use it like that - I did that yesterday with some test failures and Gemini 2.5 Flash gist.github.com/simonw/d867a...

datasette.cloudExtracting data from unstructured text and images with Datasette and GPT-4 Turbo - Datasette CloudClean data, with well defined columns and rows, is a beautiful thing - the ideal starting point for any data analysis or visualization project. Sadly, very little of the world's interesting data is pu...

By far the most interesting application like this in my opinion is structured data extraction - feeding in unstructured text or janky scanned PDF files and using the LLM to pull out "clean" structured table data www.datasette.cloud/blog/2024/da...

Valentine

@valentine.zip

dkriesel.comXerox scanners/photocopiers randomly alter numbers in scanned documentsXerox scanners/photocopiers randomly alter numbers in scanned documents Please see the „condensed time line“ section (the next one) for a time line of how the Xerox saga unfolded. It for example depi...

To me this is always equal parts amazing and terrifying because it reminds me of this talk (the German one at least): www.dkriesel.com/en/blog/2013... “Never trust a scan you haven’t faked yourself” gets a lot more scary when a scan is not as easily recognizable as such.

Generally though we've been wanting to build software that "understands" human language for decades - LLMs do a good enough imitation of that to be genuinely useful, for plugging together all sorts of tools and processes where a human can tell the computer what they want to do

TQ White II (he/him)

@tq.tqwhite.com

My related fave example happened to me the other day. I had two databases with different column names for the same data (don't ask!). I gave Claude examples of each and asked for mapping tables. Done. Correct. Seconds.

huevosabio.bsky.social

@huevosabio.bsky.social

Yes, in my experience their best use cases are for "translations" where the input has most of the information and it needs to be augmented or "translated" to another mode. In my view, transcription, natural language translation, code gen are all versions of this.