red_admiral

4 ms·

31.
by red_admiral 2mo ago
Haven't paid video streaming services solved that one already? Admittedly not with age verification, but you can set up child accounts. I presume Disney+ is ok here as they want to be in the family-friendly market but also offer conten
32.
by red_admiral 2mo ago
How many proper cross-platform approaches are there these days? Just about every one I've seen is "grab the native platform's canvas, draw your own".
33.
by red_admiral 2mo ago
Irony: the screenshot with the openclaw logo at the top lists as the first feature "Clears your inbox". What happened to write-only backups in case of ransomware?
34.
by red_admiral 2mo ago
This assumes the parent knows a little bit about computers - as opposed to the child being the one that's constantly helping the parent out.
35.
by red_admiral 2mo ago
We got into this mess because people wanted Facebook to have age verification, and MZ decided to push the responsibility on to OSes. The result looks to me like Facebook will ban access from non-approved OSes like Linux. Android will still
36.
by red_admiral 2mo ago
This was classic Mac era.
37.
by red_admiral 2mo ago
> UI constantly shuffling around between updates. and then 365 came along ... an option to pull stock market data here, a LinkedIn connection there, oh look a wild copilot button appeared.
38.
by red_admiral 2mo ago
It used to annoy me that on Windows I could press ENTER to close a message box, but on Mac I had to reach for the mouse.
39.
by red_admiral 2mo ago
This is great! My first thought reading the title was this was going to be about HTTP error code 507 because something is moving data around mechanically ;)
40.
by red_admiral 2mo ago
Something I never thought would happen: ultra-conservative Moms For Liberty and progressive education campaigners agree on something. Namely "get that AI out of our schools". The Butlerian Jihad has started.
41.
by red_admiral 2mo ago
Terrorism has always meant the other guy's goons, as opposed to my freedom fighters.
42.
by red_admiral 2mo ago
> You are supposed to manually count the question marks in your raw SQL like the JDBC gods intended. String sql = "INSERT INTO users (name, email, age, status, country) " + "VALUES (?,
43.
by red_admiral 2mo ago
Is usually the right answer!
44.
by red_admiral 2mo ago
How does the (u => u.Id == 123) lambda parse/execute? The bad way would be to load all users from the DB and filter on the server, the good way is to send a WHERE clause to the DB in the first place. I ask because `Where(u => Com
45.
by red_admiral 2mo ago
This is what you get when you let your ORM loose on the database without understanding JOINs. Especially, the bit where something like 'book.author.name' that looks like a simple field dereference actually is a method call on an O
46.
by red_admiral 2mo ago
Assuming the watermark works like the upcoming AI watermark for text, then it uses the content's entropy to embed the information. An all-black image doesn't have much entropy, so it's unlikely you'd find anything.
47.
by red_admiral 2mo ago
Probably not, if it's this kind of thing: https://en.wikipedia.org/wiki/Perceptual_hashing
48.
by red_admiral 2mo ago
> every image you create *with the help of AI*. Does in fact make a difference.
49.
by red_admiral 2mo ago
Misleading title: the watermark applies to AI generated/edited images. That includes local models. Whether it applies to non-AI generated images is a question for the reverse engineers (or ironically, a suitable AI). My bet is on &qu
50.
by red_admiral 2mo ago
But the EU mandates watermarking of AI content.
51.
by red_admiral 2mo ago
That doesn't match my understanding of how the watermark works. Let's say we have your distribution of (0.985, 0.01, 0.005). The AI somehow has to sample that distribution, let's say the pre-watermark model picks as many unif
52.
by red_admiral 2mo ago
No. Anthropic's example is completing the sentence "Isaac Newton's most famous work was called Principia ..." has only one correct answer, so nothing to watermark.
53.
by red_admiral 2mo ago
This is coming dangerously close to reinventing the X (as in X11) protocol.
54.
by red_admiral 2mo ago
Watermarking only applies where the AI has a free choice ( https://www.anthropic.com/news/claude-text-watermark ).
55.
by red_admiral 2mo ago
If anyone notices degraded quality, that would imply they could break the crypto behind the watermark. For an analogy, distinguishing AES ciphertext from random bits without the key would be counted as breaking AES (the more precise stateme
56.
by red_admiral 2mo ago
Why not? It would help with a lot of potential legal issues.
57.
by red_admiral 2mo ago
And this is why companies in the EU have a business travel rule to take a freshly wiped laptop on international trips, not just a locked one with a "distress code".
58.
by red_admiral 2mo ago
Someone is defending the side for which there's clear evidence of bots, claiming the other side did too but not providing evidence. This is what you'd expect to see if the poster were a bot linked to the same PR network.
59.
by red_admiral 2mo ago
The UK system has "severe alerts", which is what the wildfire alert was, and "extreme alerts" which is a higher category. Both trigger the siren sound, but sending this in the second-highest category seems ok to me. Both
60.
by red_admiral 2mo ago
Relevant here: https://www.astralcodexten.com/p/dictator-book-club-chavez > Hugo Chavez realized that there’s no technical limit on how often you can invoke the emergency broadcast system. You can do it every day! T

More ›