Anubis is usually less obtrusive than that, though. This is the longest anubis challenge I've ever had, to the point of being absurd. Hopefully they have a genuine reason for having set the difficulty so high.
Open AI is fudging its wording by saying more than "9 million paying business users rely on ChatGPT for work, and startups, enterprises, and governments" -- IE they're blending together paying consumers with businesses. Also not setting a time frame, so no idea how many of them are recurring users.
Other source is questionable since (for starters) it relies on sampling and mentions a 10% inaccuracy (!) rate.
The 2-3% number is from two banks, based on direct payment card data.
I think i have this too (also play piano, have perfect pitch). I have an excellent musical memory, but i have to actually play a song / sing it out loud to hear it if that makes sense.
I do have the motion-centric imagination sense (i could "feel" what it would be like to do a layup) without physically moving.
No ability to generate audio, visuals, etc "on the inside".
Nothing weird or surprising - just frequent daily work in two reasonably large applications, with a lot of testing and follow-through. I've found that as codebase size goes up, token usage does too - perhaps quadratically, and a complex back end application just costs a lot to work in, probably because it needs to map out so many pathways and interdependencies.
I tend to work in 5 or so parallel streams at a time so that also multiplies token use as a function of time. Much less and I'm waiting for it too much, much more and I can't juggle effectively. My new big project is of course to try to take myself out of the equation further and increase the parallel streams dramatically; pretty hard to get that right so still manual for now.
Well you still need land and a bunch of permits, access to materials , trucks, equipment etc. But yeah I guess big construction companies will start to automate in the next decades.
It's more interesting to me to build an entire subarb/city in a few months instead of yeras to deal with the rising housing costs. Looks like camp deflation was right but we'll see...
Credit card payments have an "offline" mode for the same reason - before ubiquitous broadband, a shop may have to dial-up to the bank to verify a card, which is slow. Debit cards that run on the credit networks can have that enabled or disabled depending on trust.
Remember originally credit card payments worked by putting the card under a slip of carbon-copy paper and rubbing off the raised digits. All the digital stuff ontop of that was just for efficiency.
I remember the stored-value transit cards in Sweden in the 90's/00's. They were just magstripe cards with a basic checksum value. It was not hard to figure out how to rewrite the card with a higher value. They'd download their transaction logs out of the buses every night and do a reconciliation. If they found a card where the value went up, they'd look at the travel dates and wait around on the "hackers" commute route with a police officer until they showed up. Very manual.
It will be possible to install the official GrapheneOS on the devices. Shipping devices with the official GrapheneOS is planned but may not be there at launch.
Not in this case. I wish I could find the actual post, but I recall reading a post on HN recently where a majority of the commenters were claiming that when they even see a Cloudflare verification checkbox that they leave the website.
This makes no sense to me as in my experience, you click the checkbox and then it verifies you without extra steps.
It can be. Sometimes you take a profile and there's a big smoking gun and nothing else matters.
Sometimes it's a lot of small things everywhere and you can pick up significant performance after a lot of small value fixes. In this case, caching wire data instead of structured data is almost one of these, because the contribution to response time for serving a cache hit is small... otoh it happens so often than a small improvement matters; but this is a pretty focused use case, you usually hit the many smalln improvement issue in a less focused application where there are many code paths.
Sometimes the whole code structure / data structures are so wrong, but it works and perf is bad and profiling will never tell you. This article is not that case; these data structures only needed refinement.
The second one is pretty much directly a Die Hard parody – new building, party, execs off doing drugs, dead guy in the lift. I'm looking forward to that one, I wouldn't be surprised if there's a lot more in there.
It isn't due to the Play Integrity API. That shows a notification on GrapheneOS with a toggle for blocking it to work around it for services not enforcing providing a result. If that was the issue, the original poster would have known from the notification. The issue ended up being PayPal shipping incorrect anti-tampering code incompatible with secure spawning. The original poster figured that out and got it working by disabling the per-app secure spawning toggle.
Tracker tools matter more than the press gives credit for. The patient apps (https://glp1.app/) capture dose timing, side effects, and injection site rotation that EMRs miss.
No, that's not correct for any reasonable definition of "impossible." Look up pangram's accuracy ratings. It's not perfect, but it's pretty good. LLMs in fact leave very distinguishing traces of their logit distributions in the text they write. It's one of the reasons why it's so easy for humans to also smell them.
It is possible to trick pangram - they bias toward a low false positive and a higher false negative - but it is not true that it is essentially random.
To put it another way: Google search works by making their own copy of the Internet. It's much smaller scale, but similarly, atproto makes downloading the entire social network relatively easy. Many ordinary techies have the resources to replicate the whole thing if they wanted to. To query your replica, you use a database, designed however you like.
So the question is, are you okay with giving random people a mirror of your public posts? After all, they're public. It's like putting them in a repo on Github for anyone to clone.
Google's Go Module Mirror is a similar but more specialized service that has a copy of all the Go modules that anyone has published on the Internet.
