> LLMs are getting very good at this and surpassing what humans have been able to prove.
But if the reverse were true it would be impossible to prove. AI is trained on all of humanities solutions - LLMs stand on the shoulders of giants, not the other way around.
As of today, the training data the originates from humans represents a vanishingly small portion of the total training. The vast majority of training - how many 9’s is a trade secret - of a frontier model is synthetic RLVR data. That’s how they got good at this.
It’s an increasing portion of training month by month. As of this last summer, pretty much all new GPU hardware is going to a combination of inference and post-training model tuning, generally RLVR.
The answer to many problems is reducing energy usage. Progress often comes from using less energy, whether that’s jamming circuits closer together on a silicon wafer or swapping incandescent bulbs for LEDs. More energy spent doesn’t guarantee a better result.
Americentrism? Code compliant where? I tried my place in New Zealand, but the address isn't recognized. I couldn't find anywhere on the site what countries/regions are accepted.
To be fair it's going to be nontrivial to implement all the codes everywhere. In some cases it's going to be impossible for an app like this to judge - In the UK, for example, there are "permitted development" rights which allow certain things (like a garden office or other outbuilding) without going through a bureaucratic process. But while some of the rules are easy to automate (eg, must be less than 2.5m high) some are not. For example, the UK allows building over 50% of the garden[1]. But if it was just measurement based, you'd be able to build over 50%, then 50% of the remaining 50%, etc etc. So the rule is "50% of the original garden". That's going to be difficult for an app like this to find out.
What province/state/territory in NZ? I can add it in. We just research the code compliance stuff manually so we couldn't cover the entire world but I can add your case in.
From my understanding this is the next iteration of OpenX4 E-Paper Community SDK - an sdk made by some of the core contributors of CrossPoint that abstracts the hardware of eink devices.
Huge props to the devs, that haven't just embraced people forking their repo - but have actually put in a lot effort to make it easier for anyone that wants to develop on these platforms.
2) pretty much just opus
3) claude code - but with Orca + app (being able to dev remotely via the app has been great for personal projects)
Work has me stuck with Cursor, which I don't find as good, even when using claude models.
reply