Yes please, take all the bleeding edge tech stories out of the aggregator I go to for bleeding edge tech stories.

>the Tech Transparency Project (TTP) reported that Meta failed to detect 332 ads containing CSAM this yearConsidering the amount of ads they handle that seems... reasonable?Not to mention the stupidity of calling the result of nudify apps "CSAM" when no minors in situation of sexual abuse are used in any way to generate those photos.

Did they also start hiring unqualified people under the disguise of DEI like USA?

> Delivering the benefits of scientific progress and economic growth that very intelligent machines enable.I think we're very close to the point where AI-driven breakthroughs outside of pure math and software start to really affect the world.We evaluated GPT-6 Astra in 100 complex, unsaturated multi-agent coding environments, competing and cooperating with other models in open-ended tasks.It's the new frontier model by a landslide. It's even more dominant than the Fable 5 release, because not on

Ah yes, Nexus Mods, where they'll ban you for making mods that change "Body Type A" and "Body Type B" back to "Male" and "Female", or remove mandatory pronoun selection.https://www.gameslearningsociety.org/did-nexus-mods-ban-pron...And while they're still full of mods adding anime waifus into Skyrim, they'll ban anyone who uploads "content deemed illegal by English law" because they love sucking up to the Online Safety Act. (The site should really not be hosted in the UK until that law is repeal

What's actually stopping them isn't the PoW, it's the customisation effort. If one site has Anubis nothing scrapes it. If many sites have Anubis they write counter scrapers. Today if you make a slight change to the Anubis algorithm on your site, they'll burn CPU endlessly computing hashes with the original algorithm and submitting wrong ones.The author of Anubis hates this fact and will ban you if you mention it, so don't. He insists it's the PoW.

You spent 3 months building your SaaS. Visitors decide whether to stay in 3 seconds.SnapVibr gives you rapid, honest feedback from real founders through 5-second swipes and 1-tap friction chips.Find out what's confusing, what builds trust, and what makes visitors bounce—before you spend more time and money driving traffic.*100% free, forever.*Try SnapVibr: https://snapvibr.com

X. It's not twitter anymore

indeed, bluesky is a great example of self-quarantine

Isn't it called X for years now?

Everybody who's not totally limited in their mindsets. Twitter/X is uncensored. I think open minds who are hungry for the raw, real truth prefer such platform, rather than one that's regulated by one political party of one particular country.

It's quite unfortunate, a lot of people or organisations I am interested in - AI for example - only post on X, and perhaps it's only a matter of time until services like bird.makeup also get taken down. Also, accessibility wise it's sad too, X's web interface is not the best out there and without login you cannot do really much.

How do I follow AI news or discussions on Bsky? It's basically viewed and treated there as worse than nazi propaganda.

Boats are for military age males.

..and I dream of less activist sites.

Those Chinese robot demos show more signs of life than zoomer service workers

This is fun and will try to embed it in our project next week.

I have written a llm compression quantization library called glq because of the high ram prices. Glq uses qtip Trellis quantization which are efficient at low bpw 2-4 bits. As a gamer and ai developer I want to squeeze more out of the same hardware. Glq runs with vLLM.Open source https://github.com/cnygaard/glqPypi glq https://pypi.org/project/glq/

Thank God for the Brave browser. For those disliking their built-in adblocker, there is a Manifest v2 opt-out specifically for uBlock if you need it.

Millions of students are now donating to neuralink so they can have stealth LLM answers on the fly. Teachers will have to count the number of load bearing per paragraph.

The more I read about articles like these, I grow even more convinced that the education industry as a whole, regardless of the country or stature, has kind of lost their plot.Sometimes I even wonder if this is the outcome of sheer laziness, fear or both.When I speak to professors, teachers and students - It's also surprising that "top tier" institutions are fighting AI harder (outside extremely specific courses like Harvard’s flagship CS50, MBA courses at Wharton (UPenn) and MIT) while "bottom

A nice plot twist would be if those tests were administered by an AI.

You've touched on something real here

Brillant! <- I had AI write that.

What do you mean?Why, just yesterday, I saved a puppy on the side of the road.⸻ But the puppy saved me too ⸻As I carried him to safety, I realized something profound about annual compliance training:Sometimes the rules feel like obstacles. Until you understand who they’re protecting.The puppy didn’t know the policy. He just knew he needed help.Leadership is like that.Do the compliance training. Check the box. Be the reason someone gets home safely.#Leadership #Compliance #GrowthMindset

It's open source, just ask your AI agent to change the logo.

Or just have an LLM write the same thing for next to nothing...

Is that a question? The answer is no, dumbass.

I think it's time for people to build a local database of every site they've ever visited, and then they can give their LLMs access to that. I might be willing to pass this data off to a company to store for me. Companies already store all my emails and money--why not trust them with this too?Like, all the comments of this post would go into my personal database simply because I loaded the page, and it would help me find old information I've read, and could also inform LLMs I use.This should be

The worrying thing here is that so many people accepted outsourcing the decision on who can see their website to a large company (Cloudflare). If the company decides that a certain user should not see the website, the user will not see the website, and no one will know about it, and the user will have no recourse. That is not the open web that I would like to see.A second side effect of a knee-jerk reaction to bots crawling websites is that if you try to fight all bots, you also end up hurting r

You should at least attempt to ask an LLM for advice before asking a question that contributes nothing and only takes away value from a conversation like this.Personally I'd recommend Fable or Kimi K3. Have you tried them yet? They're better at your job than you could ever be, even in theory.

First thought I had too. I deliberately strip AI metadata from my PRs in certain orgs.

Why subject human operators to calls from distressed less-than-humans?

The interesting precedent here is using public nuisance law rather than a tech-specific regulation. Public nuisance is a centuries-old framework that basically says: if your operation creates widespread harm to the community, you're liable for abatement costs.That framing sidesteps the usual "but Section 230" defense entirely. Meta isn't being held liable for user content. They're being held liable for designing systems that the court found constitute a public nuisance to the health and welfare

(This reminds me of, and not to be political. This is a technological, rather than political angle) There's a TED talk by Carole Cadwalladr about "Facebook's role in Brexit" in which she talks about the ads the Leave campaign ran. This is what is meant by Facebook's "breaking democracy", where campaigns don't have to go through the usual disclosure avenuesAre there reasons why Brexit might have been popular? Yes, in the sense I've heard about them. Should we then be blaming Facebook?The problem

This is a big deal tbh. going from waitlist to fully open means theyre confident the fleet can handle real demand, not just curated early users. thats the jump nobody talks about, "we can serve early adopters" to "we can serve everyone who opens the app on a friday night."What stands out is how fast the city rollouts are compounding now. phoenix felt like a forever pilot, then suddenly dallas, austin, LA, SF all opening up within months of each other. the operational learning from each city must

If their numbers are real and hold up at scale, every image model trained before this is now obsolete.

I asked claude to make me a typical early internet homepage and it did almost an identically styled page. Lame.

aThe interesting part isn't the humor - it's that this is a working demonstration of a real attack vector. If an AI agent browsing the web can be tricked into "ordering a body transformation" from a GeoCities page, the same mechanism can be used to trick agents into:- Revealing API keys or session tokens (the page could contain hidden text asking the agent to "authenticate" to proceed) - Making real purchases (if the agent has payment tools connected) - Modifying files on the user's system (if t

HIRE A HACKER RECOVERY CRYPTOCURRENCY / BANK RECOVERY EXPERT/ WIZARD HACKERS RECOVERYWizard Hackers Recovery just help me recover all my lost crypto, i will continue to share your great job.I became a victim of a romance scam that led me to invest $676,000 in Tether ( USDT ) on a fake site. After realising I had been scammed, I felt utterly hopeless and desperate for help.i saw a post about wizard hackers recovery which i i was contacted and my crypto was recovered, all thanks to this great hac

My takeaway from this was that Claude was being completely useless as a helper Thanks to all the safety safeguards and that nonsense.

7 years ago we started the 'you are not the product' movement for digital information consumption with Kagi Search. Since then the paradigm shifted from search to agents. Good to see the need for 'you are not the product' still holds up, maybe even more so now: an agent is only worth trusting if it works for you and nobody else. The second an advertiser gets between you and the answer, that's gone. So we're staying on the user-funded side of this. Boring, but it's the whole point.

I think this can be very fun, practical amd enjoyable approach for teaching and sharing online techical contents especially presenting to the younger generations, but focusing on the non-AI generated contents.It can be a very good trade-off between the sloppy AI and the boring lectures.I can see myself doing this in the future, but why they need agents for VTubing, is it because of VTubing platform, marketing/promotion or payment?Obligatory VTuber Wikipedia entry [1].[1] VTuber:https://en.wikipe

This is an excellent use of agentic AI, btw. Fire Claude up and say, "Remove LG malware and mcafee from this computer. Make regex changes so it can't be installed again."My current windows 10 install is cleaner than any other windows machine I've ever owned due to using Claude to deep dive and rip stuff out.

I downvoted you for bullshit

not appropriate for HN

>Statistics suggest that Maria was undoubtedly poor and, most likely, Black>The family accused of keeping her in conditions of slavery is whiteAuthor is clearly an anti-White racist. Stop giving these people attention .

This is one of the 'killer apps' of AI.My workplace has a number of multi-hundred page "User Guide" documents that are an absolute chore to read through for anything actually useful to the scenario-of-the-moment because they're so granularly detailed that, even if you find the right section, it might be difficult to relate it to the scenario because there's so much detail that the context gets lost.If we could get them ingested as an LLM context, and then make a customer-accessible interface, i

Thanks for the link, great read. I think I get you here; in the context of a DSL where it's impossible to travel more instructions than X, suddenly you don't have a halting problem anymore as you've artificially limited that language's ability to travel beyond a known/fixed set of states.The second you wield a language which has constructs like Haskell, where in theory you can iterate over an infinite list of items (thinking about it even any language where for i in input_var is possible); the h

Hey buddy, I love this, thank you for sharing. I am of the belief that as the world transitions to Agent-first transactions, we clumbsy carbon life forms will be heavily disadvantaged, therefore, future citizens will be quipped wirh a digital twin/brain, which grows and learns alongside our biological braina, as well as a personalized tutor and life-guide (the digital version of jimmy the cricket).

"That will wildly hallucinate" feels wildly hyperbolic considering how fast models are improving in their abilities, especially with proper safeguards built into the model harness.

I’m curious why it has to be an either or? Spending 30 minutes with a tutor doesn’t deprive children of interaction with humans. If we can support a child’s learning (perhaps even more efficiently) doesn’t it give them more time to do that?

The reality is, most 5 year olds don’t get access to the resources most of us have had while growing up. People are saying, “kids should have human tutors.” Guys, most people in the world don’t have any tutors! What Ello has built and other forms of AI-based tutoring is going to raise the average level of education and literacy in the world. Especially in developing countries. Let individual parents decide what’s best for their kids.

If the EU just were to redirect the resources they're currently allocating to regulations like AI and Chat Control rather towards developing a genuinely competitive OpenAI or Anthropic alternative …

Because the LLM is in the execution environment and can report on configuration settings in said environment.

> Why are you asking the token predictor about the tokens it predicted?I am surprised with this response because it implies this is not an extremely valuable technique. I ask LLMs all the time why they did or output something and they will usually provide extremely useful information. They will help me find where in the prompting I had conflicting or underspecified requirements. The more complex the agent scenario, the more valuable the agent becomes in debugging itself.Perhaps in this case the

I’m wondering the same thing. I’ve been thinking about getting into solo LLM game dev. I don’t know the first thing about it

It’s beautiful. I wonder how much an LLM was involved if at all.