Do you find the models generally manage to avoid making errors when placing parts (e.g. placing parts which intersect, don't connect properly, etc) or do they rely on validation and correction cycles?

On that note- have you considered adding constraints for part availability? It would be useful to be able to create designs using only parts which are on hand.

I recall seeing a project a while ago which could take a photo of a pile of lego as input and produce models that could be made with those parts. Although I think in that case the models were sorta just random arrangements.

> I did this in order to get agentic LLMs capable of designing buildable, physical things!

I've been experimenting with FreeCAD + freecad-mcp + Claude recently and have been similarly surprised how well this works. I've produced three useful parts from Claude's designs: a 3d printed case for an OLED display + some inputs + an ESP32, and a milled aluminium hotend mount adapter for a 3d printer, and a printed pen + pressure sensor mount to use the printer as a plotter. I just provided part/model numbers plus a couple of caliper measurements Claude wasn't able to find online and went through two or three cad revisions before making the parts. All three worked on the first hardware revision.

The biggest thing I've found it seems to struggle with is understanding how the parts will be assembled. The pen mount in particular was tricky to put together, having a part which needed to fit through a slightly too small space to get to its spot. Although it's hard to blame the AI for that when I didn't notice it either I suppose.

Recently I wanted to set up a slightly complex routine involving some lights and a couple of motion sensors. It feels like magic to be able to describe the behavior I want, briefly discuss the implementation, and walk past the sensor and see it in action.

I've been playing with an LLM backed reference checker for my partner who is a lecturer. For each reference in some work she's marking an agent is dispatched to read the referenced source and check that the reference is correct / accurate / not hallucinated.

It's useful to her, but it feels like it would be a waste of effort to turn into a product, because pasting the same document into Claude/ChatGPT with a prompt like "download and read each reference etc" would work pretty much just as well.

I had similar concerns about an AI backed training generation company I interviewed with. I asked how they saw themselves competing with increasingly capable generic harnesses. I don't recall exactly what they said but the impression I got was they were so focused on competing with the big LMS players that they hadn't really considered it.

> 'call 9 1 1' -> call_contact with name '9 1 1', no required params.

> 'call ambulance' -> call_contact with name 'ambulance'.

> 'call 911' -> call_contact with name '911' and no required params.

Only the last one actually used the emergency_sos function (even though the reasoning says it used call_contact). If I were to use needle in my application, how would I improve accuracy?

All of the demo setups (smart home, robot vacuum, watch, etc) could easily have access to a bigger model running on a more capable device either locally or via the internet.

> 'sleepy time' means sleeping → start_vacuum with room 'bedroom' to start cleaning

The "DeepSeek 4 Flash grade" claim seems far fetched.

Which is it?

Both agencies used pencils, but they were problematic because the graphite could break off / float around / cause shorts.

The space pen was developed by Fisher independently of NASA. NASA bought 400 of them for $2.39 each. The Roscosmos later bought 100 for the same price.

We all have access to SOTA LLMs. If I want a "clean room" implementation of some OSS library, and I can choose between paying a third party to run a script to have AI rebuild the whole library for me and just asking Claude to generate the bits of the library I need, why would I choose to pay?

I think this argument applies to most straightforward "AI generated product" business ideas. Any dev can access a SOTA coding model for $20p/m. The value-add isn't "we used AI to do the thing fast", it's the wrapping around it.

Maybe in this case the "wrapping" is that some other company is taking on the legal risk?

As a regular 30s dude, definitely not 1% by any measure, app dating had its rough spots but generally was a good time, I experienced no bitterness.

Instead I met a bunch of interesting people and found my partner. We now own a house and are talking about kids.

The real toxicity here is the idea that women at large are somehow responsible for anyone's lack of dating success.

For anyone reading this who might be dating and feel disheartened- the hard truth is that you have two options: you can either blame the group of people you're trying to attract for having faulty preferences, or you can reflect and work on yourself and your approach. Only one of these has any chance of helping you.

One thing I do agree with you on: bitterness is extremely unattractive.