Alk­mpf­er

@simonradner Is the starting prompt shared? I imagine use subagent and compress the context to reach the end. Really interesting experiment.

Claude playing Monkey Island — about as good as I was at 14 you can watch its reason while it plays

GPT 5.4 is currently reading disassembled binaries directly to establish breakpoints and find bugs. How long ago was this sci-fi?

Claude now supports dynamic looping. If you run /loop without passing an interval, Claude will dynamically schedule the next tick based on your task. It also may directly use the Monitor tool to bypass polling altogether /loop check CI on my PR

@GuidoCrosetto "Terzium" chi è, un Suo amico del liceo che studiava poco il latino? #propagandalive

Qualcuno sta cercando di far passare il messaggio che l’Italia avrebbe deciso di sospendere l’uso delle basi agli assetti USA. Cosa semplicemente falsa, perché le basi sono attive, in uso e nulla è cambiato. Il Governo continua a fare ciò che hanno sempre fatto tutti i Governi italiani in totale aderenza agli impegni presi in Parlamento ed alla linea ribadita anche in Consiglio Supremo di Difesa in continuità con tutti i precedenti Consigli, nei decenni. Gli accordi internazionali disciplinano e distinguono con chiarezza ciò che necessita di specifica autorizzazione del Governo (per la quale si è deciso di coinvolgere sempre il Parlamento) in assenza della quale non è possibile concedere nulla e ciò che invece è considerato autorizzato tecnicamente perché ricompreso negli accordi. Un ministro deve solo farli rispettare. Terzium non datur. In ultimo voglio ribadire che non c’è alcun raffreddamento o tensione con gli USA, perché conoscono le regole che disciplinano dal 1954 la loro presenza in Italia bene come le conosciamo noi.

Last year I built a tool to help with medical literature research, searching across PubMed, ClinicalTrials/gov and Europe PMC, with AI-powered analysis and clinical synthesis. Today I decided to open-source it. Hope it can be useful to someone. github.com/Pinperepette/M…

Ready to make the switch? claude.com/import-memory

We've rolled out a new auto-memory feature. Claude now remembers what it learns across sessions — your project context, debugging patterns, preferred approaches — and recalls it later without you having to write anything down.

@RogerAlsing @andreabalducci Adobe SDK is really nice, it gives you everything, but seems impossible (at least some months ago) to know the cost. Using vision model can allucinate, it is good only when you really have complex images and you need a general description (but it is not bad)

@andreabalducci @alkampfer do you have any examples for such documents? Mine has really been aimed at business focused docs like invoices, receipts etc. So I don´t quite know how well it works on other data

I´ve built my own web PDF renderer. Now trying to make white PDFs look decent in dark-mode. Is my "Night mode" yay or nay? does it look too weird?

We’re building an LLM chip that delivers much higher throughput than any other chip while also achieving the lowest latency. We call it the MatX One. The MatX One chip is based on a splittable systolic array, which has the energy and area efficiency that large systolic arrays are famous for, while also getting high utilization on smaller matrices with flexible shapes. The chip combines the low latency of SRAM-first designs with the long-context support of HBM. These elements, plus a fresh take on numerics, deliver higher throughput on LLMs than any announced system, while simultaneously matching the latency of SRAM-first designs. Higher throughput and lower latency give you smarter and faster models for your subscription dollar. We’ve raised a $500M Series B to wrap up development and quickly scale manufacturing, with tapeout in under a year. The round was led by Jane Street, one of the most tech-savvy Wall Street firms, and Situational Awareness LP, whose founder @leopoldasch wrote the definitive memo on AGI. Participants include @sparkcapital, @danielgross and @natfriedman’s fund, @patrickc and @collision, @TriatomicCap, @HarpoonVentures, @karpathy, @dwarkesh_sp, and others. We’re also welcoming investors across the supply chain, including Marvell and Alchip. @MikeGunter_ and I started MatX because we felt that the best chip for LLMs should be designed from first principles with a deep understanding of what LLMs need and how they will evolve. We are willing to give up on small-model performance, low-volume workloads, and even ease of programming to deliver on such a chip. We’re now a 100-person team with people who think about everything from learning rate schedules, to Swing Modulo Scheduling, to guard/round/sticky bits, to blind-mated connections—all in the same building. If you’d like to help us architect, design, and deploy many generations of chips in large volume, consider joining us.

Venerdi prossimo vi attendo a Roma per la Sharpcoding Conf, parlerò di Azure AI Foundry, modelli non OpenAI per le vostre soluzioni GenAi e RAG. #AzureAI #AIModels #OpenAI Biglietto gratuito: eventbrite.it/e/biglietti-sh…

GitHub Spec-Kit + Beads is now one of my more productive workflows for pure vibe coded stuff #ai #vibecoding

@ivanfioravanti Also they still have RAM, even if I do not know how many months they have before the RAM price increase can hit them too

I think it's time to buy Apple stocks, local AI is gonna push Macs sale to the next level! 🚀

@Trenitaliano quale è la ragione per cui mi obbligate a cambiare la password e non posso fare copia/incolla sul "ripeti nuova password" signore mio, non bastavano i treni sempre in ritardo

Anthropic quietly dropped 9 new Free Claude Skills tutorials 💥 Covering Projects, Excel workflows, browsing in Chrome, file editing, app integrations, task automation, and more! Zero tech background required, so anyone can learn practical AI free and fast. 🧵↓

This is awesome. You can make AI agents learn new skills from the web with up to date data. Worth a read.

Run this agent this weekend with Claude Code. Pick a repo with a bit of history and just run: npx claude-code-templates@latest --agent=development-tools/technical-debt-manager --yes That installs the agent. Then tell Claude: “Use the technical-debt-manager agent to audit this project” You’ll get a technical debt report covering structure, patterns, and long-term risks. Simple way to audit an existing codebase without adding more tooling or processes. Link: aitmpl.com/component/agen…

Nvidia's Parakeet transcribes 6 seconds of audio in 83ms on a M3 MacBook Pro, 26x faster than lightning whisper! Local real-time voice AI is here! Try it now: github.com/huggingface/sp…

“Claude Opus 4.6 ... managing a ~50-person organization across 6 repositories. It handled both product and organizational decisions while synthesizing context across multiple domains, and it knew when to escalate to a human.” Sam wasn’t joking about being replaced by an AI CEO.

Claude Opus 4.6 is available today on claude.ai, the Claude Developer Platform, and all major cloud platforms. And within Cowork, Opus 4.6 can put all these skills to work autonomously on your behalf. Read more: anthropic.com/news/claude-op…