AI on μ tamarisk.it

Generative AI and diagnosis: an anecdote

Sun, 04 Oct 2026 00:15:25 +0100

This is not the first anecdote¹ I’ve read about how AI can help support people with complex medical issues solve them, though it may be the most dramatic. Our AI Midwife is the tale of a couple struggling with infertility who discovered, after an MRI conducted and analysed by humans, that the issue was caused by a large, previously undetected, fibroid that happened to be right in prime implantation territory.

This isn’t as surprising as it might seem. Dr Google was problematic partly because the data returned wasn’t personalised. LLMs have injested a heck of a lot of information, if you hand them a bunch of data about what’s wrong and get it to analyse it (something computers are really good at) they’re statistically likely to produce the right answer occasionally.

LLMs have undeniably had a positive impact for this couple, and yet the author is still apologetic about using them. He is right: much “of the discourse around AI has become toxic” to the point where people using it thoughtfully feel they have to justify it.

This technology has a lot of problems but it’s not unambiguously bad. Like every other technology the outcome is has more to do with how it is used then any properties inherent in it.²

the plural of anacdote isn’t data ↩︎
I am aware that this is dangerously close to the gun lobby’s “guns don’t kill people, people kill people”. There is a debate to be had around access, ownership, guardrails etc. But we can’t have it if we refuse to accept that people are using it and that the outcomes aren’t always awful. ↩︎

They are stealing our stuff

Fri, 02 Oct 2026 06:48:09 +0100

That LLMs are trained on proprietary information isn’t news. That they are not as capable as the creators marketing departments would have you believe shouldn’t come as a surprise. Unless you explicitly turn it off (and probably even if you explicitly turn it off) the models injest all the information you give them and add it to their stores.

Tristan Buckmaster, a math professor at New York University’s Courant Institute, published a bombshell statement suggesting that OpenAI had stolen other people’s work and improperly attributed it.

Don’t be fooled by this summer of AI hype

So this quote isn’t on its own surprising and if you want to read the statement it’s linked from the article. What makes it interesting to me is how much it reminds me of the behaviour of “the company” in Martha Wells dystopian future.

[Murderbot] was recording all their conversations all the time, […] the company would access all those recordings and data mine them for anything they could sell. No, they don’t tell people that. Yes, everyone does know it. No, there’s nothing you can do about it.

(Wells, Martha. 2017. All Systems Red: The Murderbot Diaries)

We are sleep walking into this and other dysfunctional futures and we really need not to be. Also that’s a really good book you should read it.

AI is causing a rise in self censorship

Wed, 30 Sep 2026 17:19:37 +0100

This is the thing that people seem to have forgotten (or maybe never knew) about “AI”. It’s just averages. It gives you the most statistically likely outcome, and that is always going to lead to mediocrity.

The spread of this is the thing that’s worrying though. Human authors have been censoring themselves for a while now, removing words from their writing that are suggestive of AI authorship regardless of whether those words are ones they would have used freely only a couple of years ago.

The rare, specific details are the first to disappear. Round after round the unusual gets averaged away, and by the ninth round the thread is gone entirely.

Source: I can’t tell who’s teaching who anymore

Sat, 26 Sep 2026 09:23:07 +0100

This is not surprising. But damn. How I changed teaching after AI managed to do all my homework assignments

Fri, 11 Sep 2026 08:19:06 +0100

Some of the code output by this is absolutely wild: Armin Ronacher: Astra for Coding: Why Are We Doing This Again?

Mon, 10 Aug 2026 14:37:30 +0100

Interesting if predictable observation from Cal Newport On AI Coding and its Discontents

Fri, 03 Jul 2026 20:58:13 +0100

Opinion: I Was Not Allowed To Type Prompts Into ChatGPT During My Chalk Talk And This Is Discrimination

Probably satire. Probably written without the use of an LLM. Probably.

Sat, 14 Feb 2026 10:43:16 +0100

This is a signal boost more than anything:

An AI agent published a hit piece on me part 2

This stuff is worrying.

Fri, 06 Feb 2026 18:13:56 +0100

Many people have pointed out that a lot of the viral comments were in fact posted by people posing as bots

Which makes a change from bots posing as people I guess.

Moltbook was peak AI theatre

Thu, 31 Jul 2025 07:36:01 +0100

The impact of LLM’s on our lives will be both more, and less, dramatic than people realise. This is unlikely to end with the world exploding; which is not to suggest that it want be completely destroyed. For some it already is being.

People Are Being Involuntarily Committed, Jailed After Spiraling Into “ChatGPT Psychosis”