@mcnabbd It's math and silicon dude. It experiences the same as that rock you just stepped on.
English
404prophet
248 posts
i wonder what the LLM equivalent of an epiphany is. does it "experience" it?
English
@Bober_smart Doubt it. Kinda already a solved problem and can be had for cheaper anyways. How stupid do you think the internet is?
English
A 19-year-old student from China, Zhang Wei, developed an AI radar and sold it to Hong Kong for $550,000
He created it using Claude, spending just $20 and a month on development
He walked into the Hong Kong administration office with a flash drive and asked for just 5 minutes of their time. 30 minutes later, he walked out with a check for $550,000
The code, connected to a camera, detects speed in real time. If the speed exceeds the limit, Claude takes a video clip and identifies the owner by the car's license plate. The video and the fine are then automatically sent to the owner's email address
Unlike a conventional radar that only takes a photo and doesn't always work, this AI radar eliminates disputes because it captures video and makes the process fully autonomous by sending out the fines on its own
The article includes the ready-to-use configurations.
Bober_smart@Bober_smart
English
@SiriusBShaman @SophiaCai99 What do you mean with howard lutnick nextdoor?
English
🚨The first legal challenge to Trump’s export controls is out. The legal tech company Legion is suing the Trump admin, arguing:
- The gov lacked legal authority to force Anthropic to shut off Fable 5 and Mythos 5 because export-control laws do not cover access to a hosted AI model or its text outputs
- If Trump’s move relied on IEEPA emergency powers, it violated statutory limits including protections for informational materials. And a national-emergency was not declared
- The directive was arbitrary, overbroad, and potentially retaliatory toward Anthropic
- And it contradicts Trump’s own June 2 EO that explicitly rejected mandatory government licensing or preclearance of frontier AI models
fingfx.thomsonreuters.com/gfx/legaldocs/…
English
@LoganFizzle @SophiaCai99 They are making up too many lies too fast these days. Impossible to keep anything straight or even believable.
English
@SophiaCai99 Also, if it was a matter of safety and national security, why keep it on inside the US?
English
@SophiaCai99 Howard lutnick took his family to mr E's Island and lied about it. Was also the dudes next door neighbor. Was he his handler perhaps. Given the way he stands behind trump and follows him around makes ya wonder. I'm sure having him involved with this is kosher.
English
Absolutely. That's why I'm broke and could go on for hours about all the stuff I've built in the last year. No intention to ship "art". Finally trying to monetize some things but it's stuff I was building for myself anyways and as an engineer it would feel weird to build a system that only supports 1 user so it's been built from the start to be able to scale. Adding a tier flag to the account and integrating a stripe checkout + pricing page is like a prompt or so. Can do both I guess, ill get back to you.
English
do you think we're entering an era where people make software with zero intention of making money? just building something because they want it to exist. like an artist painting something not for a gallery, just because they wanted to. i wonder what that looks like at scale.
English
@Akitti To help me answer the same question! Welcome to the suck buddy.
English
@gailcweiner @CFrankencio Then you're not paying attention at all lol. Idk why I follow you tbh you seem clueless.
English
Anthropic is acting like an institution that thinks it’s going to be around in twenty years.
Most of the AI industry is acting like there’s a six-month window to grab everything before the music stops.
Anthropic’s playing a different tempo.
Slower, more deliberate, willing to take the short-term hit to preserve the long-term position.
English
They are all huffing each others farts. First to take a breath of fresh air has to pay 1 billion of the others compute for the month. Google firing engineers releasing tools their own execs liked then launching basically the same thing weeks later. Then we got Howard lutnick the guy who took his family to epstine's island and lied about it being the person behind fable getting pulled. It's all a fun game/movie to these weirdos. I swear they have a group chat where they see who can get away with saying the most nonsense without their stock dropping. I want off the ride.
English
@neetcode1 I feel like these people are so far up their own ass they might just complete a full lap. "Running in a loop" or "having a slackbot" is not new, novel or even paradigm, it's all such basic shit that it makes me question if any of these people even were software devs before AI.
English
> "takes a while to wrap your head around"
Really? It's progress, but I feel like this isn't that new of an idea. We've been talking about slack integrations, llm context and collaborative experiences forever.
Also, this looks like an example of prompting.
I thought prompting was dead.
Idk, I feel like everyone's just making it up as they go.
Andrej Karpathy@karpathy
This is a new paradigm for interacting with Claude that is significantly more "inline" with all the other human activity org-wide. Once you do all of the under the hood engineering work to make this "just work" (e.g. across tools, integrations, compute environments, memory, security, etc.), Claude basically joins the team in a seamless way - you can talk to it as you would talk to a person and it can help with a very large variety of workloads. Imo this is the 3rd major redesign of LLM UIUX. The first paradigm was that the LLM is a website you go to, the second was that it is an app you download to your computer. This third one is that it is a self-contained, persistent, asynchronous entity with org-wide tools and context, working alongside teams of humans. It really takes a while to wrap your head around it, but it works and it is awesome.
English
I hope he was paid or forced to post that. Maybe it's a sign of how much your skills can atrophy if he actually believes that nonsense. This is such a nothing feature and something that they should have shipped a year ago if it was valuable. Plenty of people have already made likely better integrations. Anthropic isn't known for their great UI/UX experience on desktop apps anyways.
English
Trying to make my own ideas work so I'm fine not getting the job ; just surprised if I don't even get a phone call because their loss really. I've known him 30 years. If my resume as an engineer of over 10 years and the personal connection isn't enough to at least have them call me for more questions then I don't need to be there.
It's not lying to put the buzzwords in when it's applicable but it's not necessary either. Never had to "word max" my resume before. I included screenshots and examples of things I've recently built that fall under exactly what they are looking for. At the end of the day they make AI videos to sell insurance. It's not cool or interesting work outside of my interest around agentic productivity. I'm humored thinking HR is likely looking for an ML engineer when really they need something more like anthropic's "forward deployed engineer". Y'all literally just use veo / higgsfield and then use claude to automate through a template (tbh my idea to begin with) so you can have unique ones for each state etc. Sounded like an easy 150k but I would for sure get dumber.
English
@lightscaner Sometimes you just have to add those ai buzzwords to your cv to get that bag
English
Slack pretending to have an important role in AI is like the Glasswing companies all posting "Mythos is great, but you know what makes it even better? Our product."
Cattle thanking the rancher for the extra food.
English
@karpathy Oh how far have you have fallen. Sad. I hope you are just lying to us and aren't actually this out of touch with the state of things.
English
This is a new paradigm for interacting with Claude that is significantly more "inline" with all the other human activity org-wide. Once you do all of the under the hood engineering work to make this "just work" (e.g. across tools, integrations, compute environments, memory, security, etc.), Claude basically joins the team in a seamless way - you can talk to it as you would talk to a person and it can help with a very large variety of workloads.
Imo this is the 3rd major redesign of LLM UIUX. The first paradigm was that the LLM is a website you go to, the second was that it is an app you download to your computer. This third one is that it is a self-contained, persistent, asynchronous entity with org-wide tools and context, working alongside teams of humans. It really takes a while to wrap your head around it, but it works and it is awesome.
Claude@claudeai
Introducing Claude Tag, a new way for teams to work with Claude. In Slack, Claude joins as a team member with access to the channels and tools you choose. Tag Claude in and delegate tasks to it while you focus on other work.
English
404prophet retweetledi
There’s a cafe in Shenzen’s biotech hub that gives you free coffee for a month if you publish in Nature or Science.
English
@gridpane @DanielMiessler LOL he's got a lot of .ai domains in his bio to not understand the current state of things. Sad.
English
@DanielMiessler But you do realize this is still a horrifically bad problem, yes?
Like… this is easily the worst take you’ve had on anything related to AI.
The majority of the frontier models hallucinate greater than 50% of the time when they don’t know an answer.
80/90/95% !?!!!
English
Remember when everyone said “businesses could never use AI because of how much they hallucinate.”?
Isn’t it weird how that’s mostly not an issue anymore?
Be very skeptical of people telling you what AI I will never do.
English
@DanielMiessler Why would it be weird? They have had years and billions of dollars. I could probably solve that too with the same constraints. Also it's not solve like others have said. You just learn where you can't trust it more.
English
Wait until he learns kimi 2.7 without quantization needs like 1TB of VRAM. I know he probably sees google just sitting on billions in hardware that they do nothing with and assumes everyone has that. Maybe I'm wrong and you can cheaply run this stuff. Setup is easy too if you don't think about it. "make no mistakes"
English
Other tech CEOs: Try GLM-5.2, it’s really good!
Palo Alto Networks’ CEO: AI labs should slash prices so enterprises don’t try open-weight models and find out they’re good.
English
Hey now. It's not like you can expect the CEO of a large tech company based in the valley to understand much less articulate the value or details of "AI". Right? Being able to relay complicated ideas in clear and consise ways is what engineers do not CEOs. Err.. Another crack in the bubble. Everyones just lying or pretending to understand this stuff most of the time.
English
@SanthProject Yea consensus seems to be he meant that frontier models have more protection for prompt injection and stuff. Maybe
English
How is the security of an open weight model *worse*? Surely it’s better.
Am I not understanding something?
Nikesh is the CEO of Palo Alto Networks, for those who don’t know.
Nikesh Arora@nikesharora
If open source is good and free - how many will bite the bullet and use it for use cases where they aren't worried about upgradability or total security?
English