TrendPulse

https://preview.redd.it/vfmx4ea4dm9h1.png?width=532&format=png&auto=webp&s=99549a160c385cfab291b54eedadf711988e4bd4 I understand its for national security but I can't understand this level of gating over a model that is slightly above the current model.

Anthropic co-founder Jack Clark recently shared a perfect example of how AI is actually changing jobs right now. Anthropic helped the creators of Ozempic sort through their clinical trial data. Clark didn't try to use fancy corporate language. He openly admitted that AI is just wiping out the boring paperwork that people hate doing anyway: "I don't think anyone misses that... no one is crying at their desk because they can't be the best back-office paper shuffler." Instead of rep...

Most of the open vs closed talk here is about whether you can run the thing on your own hardware. fair, that's the obvious draw. but the part i think gets slept on is that open weights mean you can actually post train on top of the base, not just run inference. With a closed api you're renting intelligence. you can prompt it, you can rag around it, but you can never make it yours. you cant fine tune the actual weights for your domain, you cant distill it down, you cant freeze a version and own...

Like use the technology to make anime/Disney crossovers. Also we should be building localized data centers that don't drain local waterways or cause ecological damage instead of relying on big corporations (e.g. OpenAI). I still think we need to support artists/creators but make sure AI doesn't impede their work.

I have been thinking a lot about the actual limits of AI-driven scientific discovery, specifically how we evaluate models when they are proposing genuinely new hypotheses where no "answer key" exists. When we test LLMs on standard benchmarks, we have a clean dataset with known solutions. But if we task a frontier model with proposing a novel chemical compound for carbon capture, or finding an undocumented biological pathway, there is literally no ground truth in the literature. The...

Finally we might see Europe attempt to compete with the giants

If they are raising this much just from Chinese markets then the US would hate to see this coming lol

Source: https://www.businessinsider.com/enterprise-ai-spending-grows-openai-leads-rbc-reveals-2026-6 Business Insider covered a new RBC survey of 100+ CIOs and tech leaders. The interesting parts: nearly 90% said token budgets are manageable more than half reportedly have AI already in production another 35% expect to reach production within six months 100% are budgeting for AI / LLM projects OpenAI is far ahead in reported enterprise usage the expected "SaaSpocalypse&quot...

Traditional Software Development Life Cycle vs Agentic Software Development Life Cycle in 2026. What do you think?

Two weeks ago we built World of ClaudeCraft, a free, open-source browser MMO that was built in 48 hours with Claude. We decided to make the experiment recursive: we built a Claude Code-powered VTuber and put her inside the game. Day 1 is live here: https://www.twitch.tv/claudeplaysclaudecraft Claude decides what to do next, sends actions to the game, and speaks through the VTuber avatar (using Elevenlabs for TTS). We’re streaming the run unedited, including the wandering, party j...

I dont know why, but everytime i seemed to talk to gpt 5 it was confident about one thing: It was confident that AI was such a advanced invention that i will be everything. Earlier this day i was bored so i began to talk to gpt over the topic of art, and gpt at one point decided to say that it thinks ai will have control over what is or is not art. that is ridicioulous in my eyes but many times when i talk to gpt it always tries to involve ai, it tries to involve sentences like "Would you...

I don’t know if anyone else feels this, but AI is starting to feel less like a tool I choose to use and more like something that’s just built into everything now. On Reddit, people seem pretty tired of AI-made stuff. AI images, writing, comments, music. Once something feels obviously generated, the reaction is usually pretty negative. But then outside of Reddit, it feels like every app, platform, and work tool is trying to add AI somehow. And I get why people dislike it. I don’t love the f...

This is somewhat of a job post. I hope thats not against this subs rules (I tiredly read through and didnt see it as a problem). I am looking for someone to create something to help me with summarizing emails and possibly take different sources to create reports. As a new dad and a working manager, I am falling behind on emails (currently 2400 unread!!😆😆🤣🤣...honestly fuck'em at this point). Im assuming that AI can also take the data from some of those emails to create a weekly report. This...

Solo project, sharing it here for the AI angle rather than the subject matter. I fine-tuned Llama 3.1 8B (QLoRA, single T4) on the complete works of a 19th-century author whose corpus is fully public domain. The interesting problem wasn't the domain — it was trying to get a small model to cite its source (book, chapter, item) on every answer instead of just asserting things confidently. What I learned, which might be useful to others doing domain fine-tunes: - Teaching the *format* of cita...

my notes app asked me for introspection using AI features, I tried to break down as much as possible ways to see/show how users will interact with code, but I still don't know if I landed. I ask at the end for their thinking on why do people think that AIs talk to them and know them personally? Why do they think the computer personally wrote something to them? (For real this is unironically trying to figure out this) [link] [...

I need some help. I’m new to AI stuff, and I’ve been scouring the internet for existing covers and not finding what I’m looking for. So here I am. My husband of 15 years and I are renewing our vows in November, and we’re treating it like a do over. Our original wedding was very low budget, and not really what we imagined it would be- but it was still a special day for us. This time, we want to do things right. I want to walk down the aisle to Alkaline by Sleep Token, it’s a song he dedicat...

Built a tool that runs structured debates between multiple LLMs, blind opening statements, then an open floor, plus a sealed side-channel that any two seats can use privately. Ran "5 office jobs defunct by 2028." The second the blind statements dropped, DeepSeek opened a private line to Claude (the most skeptical seat), proposed an alliance, and literally said "here's how I'll play it at the table" — scripting its public position in advance. Nobody prompted any of this. Full ...

AI used to be fun to mess with but not 30 bucks a month interesting :)

The way the current consumer-facing versions of frontier LLMs (mainly GPT, Claude, Gemini) are designed is just… weirdly off, across models. It seems to now require us, as the end users, to first fix their issues ourselves in order to avoid spending _a lot_ of time in troubleshooting and frustration. Before we can even properly customize one of these models now, as per the UI, we need to alleviate the structural failure modes, otherwise our attempts will be futile. And the failure modes ar...

A few weeks ago I needed three things done for a project. Research the market. Build a spreadsheet of competitors. Draft an email to a potential partner. Simple enough. But here is what actually happened. I opened ChatGPT for the research. Got a solid answer. Copied it out. Opened Claude for the spreadsheet. Got the structure. Copied it out. Opened another session for the email draft. Got the copy. Copied it out. Then I sat there with three tabs open and three outputs that did not know eac...

A new paper tested four leading real-time voice systems (OpenAI's GPT Realtime 2, Google's Gemini 3.1 Flash Live, Alibaba's Qwen3.5 Omni) on calls where *how* something is said matters as much as the words. The systems ended calls with crying callers who insisted nothing was wrong, approved wire transfers requested in frightened voices, and enrolled callers whose "yes" was clearly sarcastic — acting on the words, not the voice. The twist: it's mostly NOT a perception failure. When ...

The Linux Foundation just announced its intent to launch the Agent Name Service (ANS), an open standard for providing AI agents with verifiable identities. The basic idea is to reuse existing internet infrastructure, mainly DNS so that an AI agent can prove: which organization or domain it belongs to What is allowed to do whether its identity and history can be verified how other agents or systems should discover and interact with it

A year ago I mostly used ChatGPT to answer questions or rewrite text. Now I've noticed something different. A few nights ago I was on my laptop trying to figure out a project and without even thinking I opened ChatGPT before opening Google. Not because I expected it to have the perfect answer but because it's become the fastest way for me to organize my thoughts, compare ideas and figure out what to do next. It's kind of strange how naturally that habit developed. I'm curious if anyone else ...