

Ok, thanks for the detailed explanation. I guess if your goal is to make your model sound like another model that makes perfect sense.


Ok, thanks for the detailed explanation. I guess if your goal is to make your model sound like another model that makes perfect sense.


I wonder if you’d be allowed to read some other passages in addition to the required ones?
Like the part where the one prophet has God summon a bear from the woods to maul like one or two dozen kids because they called him bald.
Or the part where the Israelites convince another group of people that they want to make peace and convince them to get circumcized as a sign of good will, then they massacre them all while the men are recuperating from the circumcisions, and God is basically like “good one!”
There are a lot of illuminating passages that would help kids make a proper judgment as to whether the Bible is really the product of an omniscient, benevolent being.


The article is not clear on what a “distillation attack” is… what exactly is Alibaba supposed to be getting away with here? The article mentions using many different connections through obfuscation networks and proxies… so that would get them around rate limiting, and maybe enable them to submit many queries on free accounts… just spin up a new account whenever you hit the token limit of an unpaid account. So basically it’s a terms of service violation?
I don’t see why it’s necessarily a huge leg up for a competitor… they are just using the outputs of another model as training data. They still need to train their model, which is the expensive and energy intensive part.
It sounds to me like Anthropic just wants the US Government to help enforce its TOS internationally and force Alibaba to pay for those precious tokens? Because apart from that piece, the “attack” just seems like normal use of the service. If Anthropic’s service has an inherent vulnerability, that’s their problem.
Of course all the other comments about how they stole all their training data in the first place are spot on.


No no really it’s about protecting girls, really it is! Now go on over to the doctor’s office and take off your pants.


90% of MAGA would support Trump taking a dump directly on their faces.


Nope, now (by default) it opens all the files you had open the last time you used Notepad. You can turn it off, but it’s annoying.


Sure, I have no doubt that a terminated NIH scientist will have no trouble finding a “factory job” that pays two or three times as much.
What world does this guy live in?


I’m not surprised that Hannity finds it “almost impossible to comprehend” that someone would resign to uphold their principles.


Also ironic since in this case the President and Secretary of Homeland Security immediately came out stating that the victim was a domestic terrorist and tried to murder the ICE agent with her car before any investigation was conducted, and both of which turned out to be false.


Very true, and a big part of the reason most people won’t want this.
It’s like the “AI pin you wear that absorbs literally everything you see and hear” product that was pitched a couple of months ago. Kill it with fire!


The article keeps using the phrase “upcoming XR revolution” but I don’t see this gaining much traction outside purpose-built implementations for specific jobs, and a subset of tech enthusiasts.
Long-time Linux user, have never run AV on my Linux machines.
A few years back, I was forced by compliance rules at work to install AV on a Linux server and started looking for solutions. I shopped around a bit and what I found was that even the commercial AV vendors who supported Linux had no more than 4 or 5 actual signatures to detect Linux malware, and they were all 5 or more years old.
Things may have changed since then, but this may be a good way to think about it… how much Linux malware can these tools actually detect?
Yes, Linux rootkits are a thing but if your AV doesn’t detect them, there’s no point running it.
Think I like this even better than OP’s


OMG yes… I wrote a macro that copies thousands of rows and then closes a file and I had to add a step to copy just one cell before closing to work around this stupid message.


I love how many people rushed over to help her. Good citizens looking out.


I think the point you are making is that “AI is useful,” which is slightly different than “AI is great.”
I’m an AI hater, but I’ve found some niche areas in my life where it’s useful.
But I still see the overall net effect of AI - well, at least of LLMs, which are a small subset of AI - on the world to be overwhelmingly negative to the extent that the niche positive impacts are not remotely worth the costs.
Something looks kind of off when he folds in the side mirror… it seems to disappear entirely in an instant.
Otherwise, the video looks real enough… Who knows.
EDIT: on further review, the camera shifts right to an angle where you can no longer see the side of the car… and you can still barely catch a glimpse of the mirror a second or two later. False alarm.


Just when you think the US Congress could stoop no lower than Marjorie Taylor Greene, Nancy Mace steps up to the plate and hits it way out of the park.

Murder one person, you go to jail
Murder countless millions and you’re a champion of industry cutting through needless red tape.
Just remind them that it’s natural to be curious about crab literacy, but they need to find a more healthy outlet for these feelings.