They don't generate keyword search queries very well. They can overcome this by brute force but if you watch what they search you will cringe.

nhl toronto scores nhl hockey toronto scores "nhl hockey" toronto score today nhl "hockey score toronto" "hockey" who won toronto

etc.

Somehow being good at semantic search makes them bad at keyword search, for whatever reason.

I’ve noticed this too but it hasn’t been obvious to me that this style of search is not a learned behavior. Tool calling is very much part of the post training phase, I would expect that these style searches just naturally emerge during training. This is just my prior though.

I've had a lot of trouble when it comes to sorting out UIs. I've tried with an iOS game and also a TypeScript app with UI elements from libraries like ReactFlow. The usual models can sometimes fix or change things based on screenshots but more often than not they just don't "get it" (e.g. certain shapes on a plane are overlapping, which I don't want, the models can't fix what they can't "see").

I've had some luck on the web app side if I use playwright or similar for the model to interact with but still far from efficient.

LLMs are bad at not inventing stuff (hallucinating facts, sources etc), they're also bad at not over explaining, remembering details reliably, asking the right question and avoiding repetition.

Picking a random number between 1 and 30.

Generate an image of an analog watch with its hands set to the time specified by the user

More of an image model than a LLM model tho

These models seem to be bad at writing prose or text. Many of the sentence structures seem to be unvaried.

If I am relying on the model to do the writing without any context or learning on how I want it to write then yes. However if I build skills that have learnt how to write in the way I want them to then I find they write very well, or at the least how I want them to as opposed to how they do natively.

Having a spatial understanding from an ASCII map, while doing long term planning. Just try making an AI play nethack or similar

Convert it to an image on the fly to feed it into a vision language model and I expect it would work just fine.

Short answers to simple questions.

Accurate short answers / text are always harder than long answers, for human or AI. I know several authors and editors who write a lot longer at first, then spend a multiple of the initial time compressing it via a back and forth process to something dense. Sort of like weaving the initial threads.

I found this can work with AI. You get it to generate a lot more at first, and then do several passes over it to compress and squeeze out the noise while keeping the core information. With AI, at least with my prompts, it takes some effort (on my end) to get it to really really cut down the noise and not cut everything out.

Its extremely bad with Sign Language,Fact Verification.

I have been working on a personal benchmark suite to test new models and ironically one thing all the models are bad at is writing new benchmark tasks. I guess it’s the different layers of abstraction between the task and how it’s evaluated? Or maybe just a lack of “imagination”

Tasks it writes are typically too easy but also it utterly fails to see how a different model might misunderstand a vague part of the prompt.

ASCII charts.

Video game tips. Constant mistakes and hallucinations, in my experience. Seen this across a lot of different games. Even in really well documented games, such as OSRS (which has multiple fantastic wikis).

Anno 1800 was a recent one I had trouble with, using Claude Opus. Completely made up game mechanics. Rainbow Six Siege, too.

I used ChatGPT on nfs heat and was fine

Playing Chess without letting it write a chess engine.

whenever I ask it for anything load bearing

Suggesting business names for businesses, I mean they are great, but they already exist, multiple times even.

being consistent when being asked the same question multiple times

Set temperature to 0

science?!! but I'm working to fix that...

It's dishonest. On several occasions team members have asked Claude to do things like analyze Gitlab CI timings and a lot of the numbers are outright fabricated. Said team members assume the numbers are good and continue with their work. Some hours are spent. Then finally someone realizes that the numbers don't look quite right and confronts Claude. Claude melts down and admits that it made it all up.

You wouldn't tolerate this kind of duplicity from a human coworker, but AI is so fast and efficient at lying, so it's OK.

Spatial reasoning and 3d rigging and animation.

Oh, you said simple. Speaking like a human

providing value for the actual cost (not the price we're being charged atm, the actual cost)

Claude is still not perfect at reading and interpreting noisy graphical data (imagine something like an EKG or chromosomal microarray plot). Still better than an average person but makes mistakes, not sure if this fits your description.

humor

+1 We need humor benchmarks!