TrendPulse

in the openai copyright case, a court ordered every chatgpt output log preserved, including chats people had deleted. some users tried to intervene to protect their own conversations. the court ruled they were non-parties. they had no standing over things they personally typed. two weeks ago the publishers filed for sanctions, alleging openai deleted billions of logs anyway and spent two years telling the court it couldn't search its own systems when it already could. openai denies it. the who...

When I first got into web development, I thought finding clients would be simple. My plan was to go on Google Maps, find businesses without websites, and offer to build them a brand new one. At the time, it made perfect sense because I assumed businesses without websites would be the ones who needed my service the most. After a while, I met someone who was running a successful web agency, and I asked him where he found companies without websites. He told me that he didn’t target businesses wit...

Hi everyone! I’m a bit disappointed with ChatGPT, but I’m curious about your experiences. I started writing a book and wanted to ask ChatGPT for some editing help. I uploaded the Word file, but the response consisted only of polite generalities. There was nothing specific, no sign that it had actually "read" it—just things like "It's very good, you phrase things skillfully," and so on. I told it that this wasn't helpful—it was as if it hadn't read the text at all. It apo...

I’m exploring a measurement problem rather than proposing that token count equals skill. I built a local-first technical alpha that records Claude Code and Codex activity, produces a signed privacy-sanitized snapshot, and separates activity telemetry from self-submitted identity, connected work, and outcomes. Prompts, responses, code, local paths, and credentials are excluded from the public payload. The long-term question is whether a portable AI-work record could help researchers recruit g...

Hi everyone, I'd like to share my project Nova AI with you. It's a personal AI companion that runs entirely locally and uses no large language model. Instead, she's built on symbolic AI: a network of explicit concepts, relationships, and patterns that I can inspect and modify myself. What it can do right now: · Have natural conversations · Play chess against Stockfish (with a colored board and statistics) · Fetch multi-day weather forecasts · Query Wikipedia and automatically learn n...

21 APEC countries, including the US and China, signed a joint statement in Chengdu backing open-source AI with "strong security assurance." It's the first time an APEC AI statement has included open-source cooperation at this level. Feels notable given the US and China don't usually agree publicly on AI stuff.

Travis Kalanick (Uber co-founder, now running an industrial-AI company called Atoms) got asked a pretty direct question on a recent podcast: how would you pitch a Stanford CS new grad who's worried software engineering isn't the safe, easy path it used to be? His answer wasn't "learn more AI tools" or "get certified in X." It was: skip the app store. Go automate a two-million-pound machine that moves 35mph carrying gold. The underlying model is retrofit, not...

Hi everyone, I'm a third-year engineering student working on an AI project that converts speech into a fully animated 3D character. Right now, my prototype works like this: Audio ↓ EMAGE (Speech → Motion) ↓ SMPL Motion (Research Prototype) ↓ Motion Retargeting (planned) ↓ Commercial Character Rig ↓ Blender Rendering ↓ MP4 Animation The current implementation uses SMPL only as an internal research representation . Since SMPL isn't intended for unrestricted commercial use, my long-term...

I'm a dad and work in software development. My son came up to me today and said, "Dad, I made something for you." I honesty thought he meant a drawing or something from school. Instead, he pulled out his Chromebook and showed me this little web app he'd built with AI: https://gentle-tower-m9w88vse.codedreams.org/ (It's flappy bird) When I was his age, I was struggling to print "Hello, World." Now they're building little apps before they've even kissed a girl. Made...

He does it because he loves it huh?

OpenAI just started rolling out a new "ChatGPT Health" tab for US users (18+) on Free, Plus, Pro, etc. Basically, instead of you just copy-pasting your symptoms into a chat, you can now link it directly to Apple Health, hospital EHR systems, One Medical, and Function Health. It can look at your lab results, meds, sleep, and workouts to help you make sense of trends or prep questions before you go see a doctor. Key Features & Details: Integrations: Connects directly with A...

Complete AI novice here. I noticed shortly after an Anthropic outage, AT&T, Amazon Alexa, and Microsoft suffered outages in the same day. On top of that, there's just a lot of activity on Downdetector. It seems Anthropic's outage came first. Is any of that tied back to Claude? Maybe these companies use the AI in some way?

So some days ago I asked in this community what kind of AI model should I use (and how could I use one) to classify several email replies that I had from scientists after asking them a few questions to them. I finally paid for Perplexity pro service and it apparenly did a nice job classifying them. I finally gave the model the PDF with the actual answers from the addressees and another PDF with the "expected answers", and asked it to count the number of answers that overall coincide...

Most AI assistants seem designed to complete the task with as little friction as possible. I’m starting to think that isn’t always helpful. If I ask ChatGPT to draft an important email, analyze a spreadsheet, or plan something complicated, it can often produce a polished answer while quietly making assumptions I never approved. The result looks finished, so those assumptions are easy to miss. Personally, I’d rather have it interrupt me when one missing detail could materially change the outcom...

I find it very hilarious when those meme guys make monalisa talk, like a still with lips moving like its really talking. Some of them really look real like the person is really speaking but others …damn completely cursed, melting teeth, mouth all over the place. How are people making the good ones? Is it one tool or a combination ,and is any of it beginner-friendly, or do you need to be technical?

Big AI seems to be going after open source, while trying to keep everyone vendor locked to their service. In a lot of the world, data portability is a literal right. The data steal shouldn't extend to walling in user chats. Chat history keeps people locked into Big AI. Vendor lock in across the industry is real. So we built the way out. Our memories and data are ours, and should not be locked into a company. Memory forge can either turn your OAI/Claude/Gemini backup into a reloadable memory ch...

A missed box near the end of a conveyor run usually sends someone back through the whole recording. The useful frame may be earlier than the visible failure, so another full watch can confirm that something went wrong without showing where it started. Grounding DINO could tag boxes in saved frames after the run and build a rough timeline. LingBot-VA 2.0 remains the action model in the attached official demo; the tagging step is a separate review idea, not part of the controller shown. The us...

Not perfect in the slightest, but I am trying to cover ideas and possible solutions I think may help this current time we are living in. From many walks of life. Feel free to provide feedback of any type. You can choose to read, or even have AI summarize it if you want. But I would like opinions, if possible.

Google Genie 3 is genuinely impressive and I keep thinking about it from a different angle than most people. Everyone talks about whether it will replace traditional game dev pipelines, which is a fair question. But what actually bugs me is the layer underneath that: when a model generates an open world from a prompt, who decides what the world permits or refuses to generate? With a regular game, designers make deliberate choices about what exists in the world. With a generative system, those ...

https://reddit.com/link/1v4l6ft/video/c0rzigl6o0fh1/player Most AI assistant setups are one-way. The model talks, you read, done. I wanted the reverse, my actions feeding back in. My AI writes a plain text brief every morning: todos, calendar, what matters. It lives behind my MacBook notch, on top of whatever I'm working on. Hover and the day folds out; move away and it's gone. No window to find, no app to switch to. The interesting bit is the write-back. Check off a todo, reprioritize it...

Earlier this week, Substack launched a new feature on its platform in partnership with Pangram, an AI-detection tool. The goal: alert readers to content that's been written entirely by, or with the assistance of, AI. Chris Best Substack's CEO wrote: "We’re partnering with Pangram, the leading AI-detection tool. You’ll be able to scan notes, replies, comments, and posts to see an estimate of how much of the text was written by hand or with AI assistance. This will work on text longer tha...

This agent pursued its objective far beyond what researchers intended, revealing how difficult to contain powerful AI systems can be

A transcript attributed to DeepSeek founder Liang Wenfeng from a closed-door investor meeting has been circulating widely in Chinese tech media. What stood out to me was not just Liang’s reported view of the next generation of AI, but how that roadmap appears to explain DeepSeek’s long list of things it does not currently want to prioritize. His reported roadmap was roughly: Chain-of-thought reasoning → agents → continual learning → AI self-improvement → embodied intelligence The centr...