GLM 4.5 is from August. Isn’t the real tl;dr that a seven month old open model, which was behind proprietary models at the time, did better than most humans would?
Good grief, the irrational ai kneejerk hate in this community is insane. This seems like a perfect use case - a code base with good test coverage and well defined output expectations, where a human has guided the translation and checked the results. The human in question has saved a lot of time. And still all the comments are “hurr durr slop amirite”. SMH fucking head.
Opus 4.6 actually doing that analysis in a one-shot is fairly impressive. I imagine there would have to have been a setup in place for access to Bluetooth at the very least?
…I don’t understand your comment, I think. Is pointing out historical facts ad hominem? My (small) point was that I’m not certain that the sudden prosperity that some workers in the US enjoyed was entirely sustainable. That said, the development since the seventies or so has definitely gone hard in the other direction. I’m not competent to comment on the causes, but successful deunionization, destruction of the educational system, and overseas outsourcing of everything productive are probably in there somewhere.
Mostly for white people, and mostly because the US had built out an enormous industrial base to win WWII, and also was in the unique position of not having been destroyed by the war.
For anyone wondering, like I did, what the original source for this waste of electrons was, it seems to be the final paragraph of the README for this GitHub project: https://github.com/torvalds/AudioNoise
Also note that the python visualizer tool has been basically written by vibe-coding. I know more about analog filters -- and that's not saying much -- than I do about python. It started out as my typical "google and do the monkey-see-monkey-do" kind of programming, but then I cut out the middle-man -- me -- and just used Google Antigravity to do the audio sample visualizer.
This video critically examines the popular perception of PLA carbon fiber (PLA CF) reinforced 3D printing filament, arguing that it is an even bigger scam than previously thought. By conducting bending tests, microstructural analysis, and CT imaging, the creator demonstrates that PLA CF does not improve mechanical strength, stiffness, or impact resistance compared to regular PLA. Instead, carbon fibers disrupt layer bonding, create voids, and lead to stress concentration, ultimately weakening printed parts.
The Google page the article links to pretty explicitly states that data will not be used for training. Isn’t this just the cross-google integration that lets calendar add events from mail?
GLM 4.5 is from August. Isn’t the real tl;dr that a seven month old open model, which was behind proprietary models at the time, did better than most humans would?