preciousoo

I just noticed that it’s a config option. Weird that it’s so short though, I can understand why it may be needed for users who spawn hundreds of sessions a day

Isn’t “Anthropic won’t fix it” a little sensational for barely month old issue with little activity(two upvotes, one of which is me), in backlog of 5k+? Agree that it’s a real issue that need fixing however

The link sharing platform is full of people interested in sharing links, that makes sense, similar to how a mall would be full of people interested in shopping

What definition of “often” includes a 0.21% rate?

I created a watcher for this problem, to watch my PRs for unfinished scope and have a fresh Claude review

Uses tmux and gh https://github.com/Kyu/claude-pr-watch

I dont uninstall apps that annoy me with notifications, but I do disable them. Most of my notifications these days are news or texts. So be it

Or create "blind" benchmarks.

10 groups of 3 researchers, all have their own benchmarks that they do not share (testing it without the authors knowing is a different problem, maybe they only run the benchmarks when the gen-pop has access to the models).

that's 10 different tests. Aggregate pass rates

“Exam Question: {prompt}” was enough to get me the right answer on whatever model you get with logged-out ChatGPT.

Neither prompt was enough for llama3.3 or gpt-oss-120b

Pydantic + FastAPI are my two favorite python shops right now, they’re always dropping fun new projcts

The PhD can't read minds, the quality if the request from a moron would be worse than the quality of the request from someone with avg intelligence. And the output would probably noticeably differ accordingly

I guess we’ll be rebuilding society from scratch anyways

It depends on if the people with weapons can find the people with food. With no rule of law, everything is on the table. Warlords still exist in many parts of the world today

Search/Data in general is incomplete and doesnt match the heatmap. Good idea though

makes sense, I didnt know you could have multiple (learned in this thread)

I’m wary ofhardware 2fa because I’m prone to losing things. Do you have a plan for if that happens?

The take home assignment is to create a web service that changes the color of a cube, hardly anything cutting edge. Do you expect to be paid for coding interviews too?

Click the link of the take home and tell me if you expect it to take you more than 20 minutes. I don’t understand the resections here

I wonder if the people replying bothered to look at the take home question. I wish I had an interview like that, it would be the second easiest interview I’ve ever done in my life, and then task is interesting and easy to me, as someone who just started learning unity a week ago

Hi, I’d like to work for your company Great, state your price and availability If match: create a simple project(<1 hour of work) that demonstrates familiarity with technologies that will be used on the job. If done, check for team fit.

This is toxic to you???

Seems like a self fulfilling prophecy

Coming into the thread(and general discussion about chatgpt being used as evidence) with this context, I’m confused about the reactions to this. Online activity has been used as evidence as far as I remember. OpenAI also has a couple high profile cases against them with chatgpt history used as the primary evidence

classes are optional in python, does that violate the spirit?

That's ~6% of the 9M. Not nothing but not a glaring amount. I'm seeing that there are 1.7M total SWEs in the US, which would bring the H1B % to 30%, assuming all are SWEs. This would make it a significant amount, not sure if it's enough to suppress wages / pass on US persons if they attempted it.

One look at the top H1B employers should stop this narrative however, because those companies do not pay near peanuts.

https://www.uscis.gov/tools/reports-and-studies/h-1b-employe...

It’s an absurd statement because you are human and are aware of how research works on an individual level.

Take yourself outside of that, and imagine you invented earth, added an ecosystem, and some humans. Wheels were invented ~6k years ago, and “humans” have existed for ~40-300k years. We can do the same for other technologies. As a group, we are incredibly inefficient, and an outside observer would see our efforts at building societies and failing to be “brute force”

This is a multi billion dollar ecosystem with no regulators but themselves. Basically state level capacity. Guns are small time for them if they decided to enter the arms market fully. My biggest guess is that the threat of having the biggest military in the world as their neighbor is the only thing stopping them (missiles are much bigger priorities than narcotics)

You severely underestimate the capability that they possess, both intellectually and materially. The skill(and more critically, engineering environment) that went into manufacturing the boat shows a lot

They'll probably move onto mass producing weaponry, which, depending on the sophistication and scale, could be big issues for the rest of the world. They already partner with terrorist groups and other unsavory orgs as-is. Any group worth mentioning these days interfaces with the cartels

Funny how the most advanced anti cheat just gives version info and executables in one nicely human friendly package. No need for gimmicks when you the work speaks for itself

fwiw I couldn't find the endpoint in question for vanguard, but I did find for all the riot games

Thanks!

Holy shit. I’ve also been similarly creeped out by this behavior as well, and did the same setting changes as the original commenter. This comment made me remember I signed up for a Delta/Lyft promo way back in college. In my case in explicitly signed up to link both accounts, but I wonder if they have other methods