German health insurance will reimburse a mental health app within days but cannot provide a therapist within six months. Last week, psychotherapy fees were cut by 4.5%. Baumol’s cost disease — originally about why string quartets get relatively more expensive — explains why the app gold rush and the collapse of mental health provision are the same phenomenon.
There is no seahorse emoji in Unicode. Ask a large language model to produce one and watch what happens. The failure is not a hallucination in the ordinary sense — the model knows what it wants to output but cannot output it. That distinction matters.
A new video went viral last week: same question, “should I drive to the car wash?”, different wrong answer — the AI said to walk instead. This is neither the tokenisation failure from the strawberry post nor the grounding failure from the rainy-day post. It is a pragmatic inference failure: the model understood all the words and (probably) had the right world state, but assigned its response to the wrong interpretation of the question. A third and more subtle failure mode, with Grice as the theoretical handle.
A viral video this month showed an AI assistant confidently answering “should I go to the car wash today?” without knowing it was raining outside. The internet found it funny. The failure mode is real but distinct from the strawberry counting problem — this is not a representation issue, it is a grounding issue. The model understood the question perfectly. What it lacked was access to the state of the world the question was about.
AI companions promise to address the loneliness epidemic. Daniel Wegner’s ironic process theory predicts they will fail under exactly the conditions where people need them most — and recent data from MIT and OpenAI suggest the prediction is correct.
The most famous choice in science fiction is epistemically impossible to make rationally. Morpheus offers Neo ’the truth’ but gives him no way to evaluate the offer. Cypher’s decision to go back is more philosophically coherent than the films acknowledge.
The Oracle is the most interesting character in The Matrix for anyone who thinks about AI alignment. She systematically lies to Neo for his own good. The films present this as wisdom. I think it is a cautionary tale the Wachowskis didn’t know they were writing.
On 2 December 2024 I gave three workshops at HfMT Köln’s Thementag on AI and music education. The handouts covered data protection, AI tools for students, and AI in teaching. This post is the argument behind them — focused on the curriculum question that none of the tools answer on their own: what should change, and what should not?
On October 8, 2024, Hopfield and Hinton were awarded the Nobel Prize in Physics. The physics community reacted with irritation: is machine learning really physics? The irritation is wrong. The energy function of a Hopfield network is literally the Ising Hamiltonian. The lineage runs from Giorgio Parisi’s disordered iron alloys in 1979 to the model that predicted the structures of 200 million proteins.
Three natural fairness criteria for an AI classifier — calibration, equal false positive rates, equal false negative rates — cannot all hold simultaneously when base rates differ across groups. This is not an engineering failure. It is a theorem. Choosing which criterion to satisfy is a political decision, not a technical one.