Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Spatial intelligence is AI’s next frontier

drfeifei.substack.com

Post reply on HN
Print view View original

51–60 of 136 posts

Re: Spatial intelligence is AI’s next frontier

#51

Post by programjames »

storus wrote: post #30
programjames wrote: post #14

Earlier quoted context omitted.

Large latent flow models are unbiased. On the other hand, if you purely use policy optimization, RLHF will be biased towards short horizons. If you add in a value network, the value has some bias (e.g. MSE loss on the value --> Gaussian bias). Also, most RL has some adversarial loss (how do you train your preference network?), which makes the loss landscape fractal which SGD smooths incorrectly. So, basically, there'…

We might not even need RL as DPO has shown.

> if you purely use policy optimization, RLHF will be biased towards short horizons

> most RL has some adversarial loss (how do you train your preference network?), which makes the loss landscape fractal which SGD smooths incorrectly

Re: Spatial intelligence is AI’s next frontier

#52

Post by inciampati »

Just had a fantastic experience applying agentic coding to CAD. I needed to add some threads to a few blanks in a 3d print. I used computational geometry to give the agent a way to "feel" around the model. I had it convolve a sphere of the radius of the connector across the entire model. It was able to use this technique to find the precise positions of the existing ports and then add threads to them. It took a few tries to get right, but if I had the technique in mind before it would be very quick. The lesson for me is that the models need to have a way to feel. In the end, the implementation of the 3d model had to be written in code, where it's auditable. Perhaps if the agent were able to see images directly and perfectly, I never would have made this discovery.

Re: Spatial intelligence is AI’s next frontier

#53

Post by jacquesm »

inciampati wrote: post #52

Just had a fantastic experience applying agentic coding to CAD. I needed to add some threads to a few blanks in a 3d print. I used computational geometry to give the agent a way to "feel" around the model. I had it convolve a sphere of the radius of the connector across the entire model. It was able to use this technique to find the precise positions of the existing ports and then add threads to them. It took a few t…

OpenSCAD or something like it?

Re: Spatial intelligence is AI’s next frontier

#54

Post by JohnHammersley »

inciampati wrote: post #52

Just had a fantastic experience applying agentic coding to CAD. I needed to add some threads to a few blanks in a 3d print. I used computational geometry to give the agent a way to "feel" around the model. I had it convolve a sphere of the radius of the connector across the entire model. It was able to use this technique to find the precise positions of the existing ports and then add threads to them. It took a few t…

Thanks for sharing, I'm interested to know more about how you did this if you have a longer write up somewhere? (or are considering writing one!)

Re: Spatial intelligence is AI’s next frontier

#55

Post by t_mann »

inciampati wrote: post #52

Just had a fantastic experience applying agentic coding to CAD. I needed to add some threads to a few blanks in a 3d print. I used computational geometry to give the agent a way to "feel" around the model. I had it convolve a sphere of the radius of the connector across the entire model. It was able to use this technique to find the precise positions of the existing ports and then add threads to them. It took a few t…

CadQuery? Would be appreciated if you're inclined to do writeup of your lessons learned.

Re: Spatial intelligence is AI’s next frontier

#56

Post by btbuildem »

inciampati wrote: post #52

Just had a fantastic experience applying agentic coding to CAD. I needed to add some threads to a few blanks in a 3d print. I used computational geometry to give the agent a way to "feel" around the model. I had it convolve a sphere of the radius of the connector across the entire model. It was able to use this technique to find the precise positions of the existing ports and then add threads to them. It took a few t…

I'd love to hear more about this -- I'm messing around with a generative approach to 3D objects

Re: Spatial intelligence is AI’s next frontier

#57

Post by byearthithatius »

dauertewigkeit wrote: post #39
byearthithatius wrote: post #31

Earlier quoted context omitted.

Curating a dataset is vastly different than introducing a new architectural approach. ImageNet is a database. Its not like inventing the convolutions for CNNs or the LSTM or a Transformer.

CNNs and Transformers are both really simple and intuitive so I don't think there is any stroke of genius in how they were devised. Their success is due to datasets and the tooling that allowed models to be trained on large amounts of data, sufficiently fast using GPU clusters.

"CNNs and Transformers are both really simple and intuitive" and labeling a bunch of images you downloaded is not simple and intuitive? It was a team effort and I would hardly call a single dataset what drove modern ML. Most of currently deployed modern ML wasn't trained on that dataset and didn't come from models trained on it.

Re: Spatial intelligence is AI’s next frontier

#59

Post by atlex2 »

koakuma-chan wrote: post #18
7moritz7 wrote: post #11

Earlier quoted context omitted.

That has been solved with RAG, OCR-ish image encoding (deepseek recently) and just long context windows in general.

Not really. For example we still can’t get coding agents to work reliably, and I think it’s a memory problem, not a capabilities problem.

On the other hand, test-time weight updates would make model interpretability much harder.
Post reply on HN