There was an article yesterday where people were complaining about "AI smell", the author showed up in comments to state clearly and unequivocally that he didn't use AI to write it, and people wouldn't believe him.
AI didn't get its tropes and tics ex nihilo. Some people just write in a way that "smells" like AI to others.
Right?! Anyway for anyone who hasn't read it, go pick up a copy. Cat's Cradle is not too long, and it's funny and engaging and altogether worthwhile. And you'll always think of ice/9 when you see an article like this one.
I manage teams and a standup with 20+ attendees sounds like hell to me. We keep standups to team scope and 10 minutes long (20 minutes in the case of our largest team, but it almost never goes the full time).
We have some larger meetings that are closer to what you are describing, but they are for higher-level management, not line engineers.
It seems like it would have been so easy for them to test programs against the failure case in the same way @yuvadm did. Just make it part of the competition rules that if an application solves the problem with quantum calls replaced with a random number generator, then it's not demonstrating true quantum improvement and is disqualified.
They said up front in the rules that they didn't want Failing With Style approaches to win, so why didn't they explicitly test for that?
Because then all entrants would fail the competition.
There has been no significant jump in capabilities such that a meaningful demonstration of general purpose or cryptographically-relevant quantum computing could be performed on "public hardware".
Presumably the organizers know this but still have incentives to drum up news about QC. So they ignore that problem and focus on how to obscure the fact this is a dog-and-pony show.
I mean, I'm sure you're right, but it would seem like the organizers were teeing themselves up for an inevitable "gotcha" moment, given that it is pretty easy for third parties to do exactly what was done here.
Paying 1 BTC for the privilege of getting dunked on in tech media seems like a bad trade.
What level of subscription are you on? If you're complaining about running out of tokens but are happy to pay "whatever" for it, it should be Max 20x, yes? And one prompt drained all your tokens for the day on Max 20x?
I am on max 10x ($100 a month). I asked it to summarize a small codebase (2000 lines). Instead, it ran 4 agents in the background, and those agents went nuts, started reading everything and related dependencies, and sucked all my daily quota, forcing me to wait till 2 AM to continue using.
Up until last month, a $100 plan was more than enough, and it was difficult to run out of tokens per day for me. Something fundamentally changed, and Claude started making more mistakes and using more tokens. I know I am not tripping because I used it for over a year; this is absolutely new.
Agreed, and sorry for the snarky tone in my reply. The $100 / month plan is what I use at home and it has always been more than sufficient for something like what you describe. It does sound like maybe you have MCPs or skills that might be changing the default behavior as the other poster suggests, or the agent somehow interpreted your request as wanting it to read node_modules in addition to the code or something. In any event I hope you or Anthropic or both figures out what the issue was.
Maybe you need to clean up MCP tools? My codebase is very large (hundreds of files, hundreds of lines per file), and I am managing fine on a 5x subscription, chatting all day.
Some time ago I cleaned up a bunch of MCP tools I had installed some time ago and that did make a significant impact on token usage.
reply