Yeah, I have been seeing lots of comments, tweets, etc, but given everything I have learned about these models - i do not think the change to 1M was innocuous. I'm not sure what they've claimed publicly, but I'm fairly certain they must be doing additional quantization, or at minimum additional quantization of the KV cache. Plus, sequence length can change things even when not fully utilized. I had to manually re-enable the "clear context and continue" feature as well.
I used the heck out of it when it was announced, and it felt like I was using one of the best models I've ever used, but then so were all of their other customers, I don't think they accounted for such heavy load, or maybe follow up changes goofed something up, not sure. Like I said, the 1M token, for the first few days allowed me to bust out some interesting projects in one session from nothing to "oh my" in no time.
I'm thinking they should go back to all their old settings and as a user cap you at their old token limit, and ask you if you want to compact at your "soft" limit or burst for a little longer, to finish a task.
i feel like i misunderstand the UI - it seems to show coalition strikes mostly being intercepted? or does the bottom row refer to strikes against coalition forces?
I suspect a subpopulation of software development is going to become a bit religious, for a short while, split into "morally pure anti AI" and those who are busy using software as a means to an end to solve some real world problem. I think the tools will eventually be embraced, out of necessity, as they become more practically useful (being somewhere around "somewhat useful" right now).
As a result, I think we'll eventually see a mean shift from rewarding those that are "technically competent" more towards those that are "practically creative" (I assume the high end technical competence will always be safe).
if you think your code is art like mozarts music, then you're probably part of the first group rather than the group trying to simply get something practical done with software as a means to do it.
Should Mozart have constructed the instruments himself? Or plucked the strings himself? No, he had someone else take care of all that so he could compose music. AI can be used the same way: take care of boring stuff so I can compose a solution to a real world problem. No, that doesn't mean AI has to do everything for you, which outright bans don't seem to be able to comprehend.
the stacked changes support, for me, was an absolute game changer. the auto rebasing, etc, is -really- nice.
i found it especially useful for Gitops type stuff where you have to make lots of little PRs
The iPhones autocorrect is one of my biggest frustrations coming from Android a few years ago. The biggest frustration for me is the tendency to correct the _second to last word_. I have never gotten used to this. I know i can stop it by "clicking" on the word instead of hitting space - but that feels slow and bad.
There is a podcast, Memory Hole, about the recovered memory movement. There is a lengthy section about this book and the people behind it (not positive)
reply