The more apt word is steganography, rather than watermark. Basically subtly adjust the weights of the model so that some subtle patterns appear. Think of how AI text prefers certain words and phrases that ordinary humans don’t use as often, like “delve,” but presumably much more subtle.
And no, as Anthropic has already said, this watermark may not survive editing/formatting.
Claude Now Watermarks Your Text | Vanja Petreski - https://vanja.io/claude-invisible-watermark/
Chinese being more token efficient is a myth, and seems to stem from the superficial fact that characters are only visually more space efficient.
The fact that each Chinese character takes up 3 bytes (as opposed to 1 byte of English), words in Chinese typically require compounds of several characters, and that tokenizers have a limited vocabulary limited to mostly English means that Chinese is actually token inefficient.
No, Chinese Is Not More Token-Efficient Than English for LLMs | markhuang.ai - https://markhuang.ai/blog/chinese-token-myth
Damn we got fate grand order memes now
Instead of searching for existing gifs, Google will now pipe your search terms to Gemini so it will generate a gif for you.
Google Translate’s backend has been moved to Gemini since December 2025, and is vulnerable to prompt injection. Have a foreign phrase to translate, then input some meta instructions in English underneath it, and it’ll follow the possibly malicious meta instructions.
Google states that this move was to introduce more features, such as conversational mode.
Google Translate’s Gemini Mode is Vulnerable to Prompt Injection - https://winbuzzer.com/2026/02/10/google-translate-gemini-prompt-injection-vulnerability-xcxwbn/
Google Translate gets new Gemini AI translation models - https://blog.google/products-and-platforms/products/search/gemini-capabilities-translation-upgrades/
Actually, with the exception of the crown, none of the jewellery had much gold
Eh, I don’t think NPUs are ready to be marketed so heavily, but they’ve been around for a while and do get used.
They’re basically a rebranded tensor processing unit, think a more specialised GPU that’s even more energy efficient at tensor/linear algebra.
It’s mostly used in more technical applications, such as image/audio/video processing, machine learning, or really anything maths heavy. Apple’s M series had NPUs, and are an understated reason why they perform extraordinarily well in a lot of scientific applications.
Uses for consumers are not as compelling (especially on laptop/desktop), mostly faster/more efficient subtitle generation, face recognition, and maybe blurring your zoom background.
It’s only available as a -bin, which means it belongs in the bin
Play stupid games win stupid prizes
Usual suspects
Eh, kind of both.
When researchers peeked into which areas of the image were being used, it showed that the tiny camera watermark from the Google Streetview car was being used by the model a lot.
That is, the recognition system had learned all the routes every Google Street view car had taken, and was using that in its recognition process.
Not all images have this watermark though, so in the cases the watermark didn’t exist it then resorts to more traditional geoguessr tactics.
Inb4 Trump places tariffs on and sues the UN
IIRC there were some polls for how helpful LLMs were by language/professions, and data science languages/workflows consistently rated LLMs very highly. Which makes sense, because the main steps of 1) data cleaning, 2) estimation and 3) presenting results all have lots of boilerplate.
Data cleaning really just revolves around a few core functions such as filter, select, and join; joins in particular can get very complicated to keep track of for big data.
For estimation, the more complicated models all require lots of hyperparameters, all of which need to be set up (instantiated if you use an OOP implementation like Python) and looped over some validation set. Even with dedicated high level libraries like scikit, there is still a lot of boilerplate.
Presentation usually consists of visualisation and cleaning up results for tables. Professional visualisations require titles, axis labels, reformatted axis labels etc, which is 4-5 lines of boilerplate minimum. Tables are usually catted out to HTML or LaTeX, both of which are notorious for boilerplate. This isn’t even getting into fancier frontends/dashboards, which is its own can of worms.
The fact that these steps tend to be quite bespoke for every dataset also means that they couldn’t be easily automated by existing autocomplete, e.g. formatting SYS_BP to “Systolic Blood Pressure (mmHg)” for the graphs/tables.
Ironically the French are very passionate about their cheese made from raw milk.
It’s still classified as somewhat dangerous, and people in high risk categories are advised not to consume it.
11 sick in French cheese E. coli outbreak | DAIRYSAFE - https://dairy-safe.com.au/11-sick-in-french-cheese-e-coli-outbreak/
Yep, laser toner is literally plastic dust that gets rolled on then baked. That’s why it doesn’t smudge or rub off when wet - it’s plastic.
That’s why there’s so many large warnings to properly recycle toner cartridges.
Tbf as long as the toner doesn’t spill anywhere, the risk is lower than other background sources like tyre dust. Still gives me the hereby jeebies though.
How printer dust is polluting the air? – TCTEC® Innovation - https://tctecinnovation.com/blogs/daily-blog/how-printer-dust-is-polluting-the-air
Laser toner is literally micro plastic powder that gets rolled onto paper then baked, mmm
That’s even more confusing, why not forego the shower glass entirely and just have a shower curtain instead?
Yes I know it’s likely retrofitted, but I won’t pass a chance to say that shower curtains are superior. Shower glass looks glossy in brochures, but everyone IRL has a sad squeegee hanging from it.
Africa tech is just Chinese tech so they’re already cooked.
Australia basically banned encryption long ago and in doing killed off their tech industry.
We need a New Zealand tech revolution, which is where all the Australian IT firms relocated to following the encryption ban.
Encryption laws are hurting Australia’s tech sector, Atlassian says | Australian security and counter-terrorism | The Guardian - https://www.theguardian.com/australia-news/2020/jul/27/encryption-laws-are-hurting-australias-tech-sector-atlassian-says
Go nazi declare bankruptcy
Redpilled, Get killed
Go fash lose cash
Lick boot, lose loot
Go right eat shite
All incel, no braincells
Nazi salute, go destitute
ADL, common L
Ku Klux, lose bucks
Apartheid, all downside