Meir Dick

@PovilasKorop @pushpak1300 Double yay! This was a persistent issue.

@pushpak1300 Yay!

Damn, so even newest LLMs still do that. They still generate Laravel migrations at the same second timestamps. Cause they run `make:model ABC -m && make:model XYZ -m` Which works well in SQLite, so all tests pass. But if I switch to MySQL... This time, it was Opus 4.8 Medium.

Like what is this wizardry? stripe.press/scaling @stripepress

This is an absolutely beautiful site. As a self proclaimed expert in buying books that look great on your shelf. press.stripe.com

github.com/laravel/ai/rel… ✨

Personal dev stack: L1 Ghostty + Herdr + Termius ______________________________ L2 Hunk + Nvim ______________________________ L3 Pi + Opencode ______________________________ L4 pi-ext + Plannotator + pi-subagents ______________________________ L5 kimi-2-7 + GPT-5.5

@maigrateAI @benedictevans @heyfeifer Exited to be attending!

Our keynote sessions are set! The Practical AI Conference’s keynotes: @benedictevans and @heyfeifer. Not to mention other brilliant minds and some serious companies in the space. 🎟️: TurnAIintoROI.com

@thdxr @TelnyxDevs is crack. So much better than twilio, the MCP is full feature, and voice AI is great too.

i told opencode to use my browser to sign up for telnyx and give itself a number they had agent instructions for sign up! it even had this crazy bot challenge the rest of the onboarding was tailored for my agent and it got a number and got everything setup

When I’m trying to improve the user experience of my applications, one of the most valuable things is being able to see an entire user flow as a storyboard. Not just one screen or screenshot at a time. This is something I love using the `/visual-plan` skill for. You can describe any flow you want, and the agent will look through your code and wireframe out a storyboard of what the flow looks like. Then you can visualize the steps in a simplified way and spot areas to improve. Recently, I found that in certain flows we were still asking for organizations, even though I thought I had gotten rid of that and made it automatic. A quick storyboard let me see all the different code paths in a simple, visual, intuitive way. Spot the areas of the flow I didn’t want. And have the agent fix it. Sign up, onboarding, and setup flows are usually some of the most important experiences in your app. And usually the least looked at. Especially because it can be hard to reproduce every flow, for every situation, for every user type, feature flag, or whatever else you have. The `/visual-plan` skill lets you visualize any part of your code. Either to understand the current state, plan out a new state, or recap updates that were made. I’m pretty addicted to this skill. I use it for a lot of other things too, so let me know if you want to see videos on those. And of course it’s all open source. You can grab it on my GitHub. I'll link to it in the thread. If you try it, let me know your feedback.

Found a word that describes Canada. Not even funny....too relevant. Read that twice - tell me its not about Canada.

☢️You remember? Google optimized Shor's algorithm. The algorithm that breaks asymmetric cryptography (RSA, elliptic curves) once you have a quantum computer with enough Qubits. The US government blocked the paper. So Google published a Zero Knowledge proof instead: a mathematical proof that they have the result, without revealing how. Cryptographic sorcery 🧙 But the Internet is sneaky. Someone launched a contest to re-discover the result with AI. The LLM searches a huge space of circuits (each one a candidate optimization of Shor's), and tests whether it beats the previous best. The clever part: they use the ZKP verifier as the reward function. No false positives, and it turns out to be a very efficient signal. In less than 2 days, the community re-discovered Google's result !!! 🔔15 days later, the LLMs are still improving it. They're already 44% ahead of Google. Hard to say where this stops, ie. what the true minimum quantum complexity for Shor's is. But we will not close the full gap. You still need a Quantum Computer with a relatively large number of qubits. The only thing that changed is that this number drops a little every day, and it has been dropping for 15 days straight.

We are doing it. 📣 The Practical AI Conference 🗓️ Tuesday, July 28, 2026 📍 Staten Island Cancel all your meetings. Tickets now available. ⛓️‍💥: PracticalAIconference.com

imo there’s a pretty solid default recipe that everyone should use to optimize a system of Agent = Model + Harness you should “train” both 1. Build v1 agent using a sensible base harness and some task specific prompting + tools 2. Harness Engineering using eval tasks that roughly match prod this is often enough - most companies can get acceptable perf doing this. then they collect traces, mine them for patterns, and make slight tweaks from there 3. SFT using data collected from traces) or synthetic data. Often is good candidate for “distillation tasks” to train a cheaper model while maintaining existing performance 4. RL if you have the bandwidth and ability and desire to create environments and designing rewards that represents the tasks you want your agent to be good at. Push past the SFT behavior of “copying” data from existing model to pushing past in some dimension 5. Light harness engineering again to squeeze any more juice (ex: slight prompting) using the trained model that’s better at your task distribution this loop will largely be productized as a general purpose recipe for building and improving agents we’re still in the earliest innings of the world’s companies getting comfortable with steps 1-2 of this loop. Harness engineering will probably be the dominant way ppl will optimize agents but i expect a large number of companies to onboard through this entire loop on some trial project of interest in the next year

It's a crime that this repo only has under 400 stars. /autoreview and /handoff are literally two of the best agent skills I've ever used. Massive thanks @steipete and @openclaw

Canada’s not going for gold with this strategy. On the consumer side, this strategy gets a lot of things right and this government deserves credit for that. But from a digital and economic sovereignty perspective I think it falls short. What the AI industry most needs is for governments to create the free market conditions for AI companies to start, scale and thrive here. There are important elements in the strategy: sovereign compute, a public supercomputer, AI Missions starting in health care, a fund to scale Canadian champions, commercializing the Photonics Fabrication Centre, and the government as an anchor customer. Members of the Build Canada community have publicly written about many of these ideas and it's good to see these show up in the plan. But we have to be honest about what this strategy is – and what it isn't. This is primarily a strategy to help Canada use artificial intelligence, with government in a main character role in framing public perception. It is not a strategy to make Canada the best place on earth to build it. Its own organizing goal – adoption, moving Canadian businesses from 12% to 60% uptake – is necessary but not sufficient to make Canada a global AI leader. The strategy says it wants Canadian champions. But you don't build champions with government cheques and deferred studies – you build them by making Canada the best place on earth to start and scale a company. Using AI and Building AI are different goals, and they require different instincts. The companies that will define this century will get built where four things are true: 1. founders and engineers keep what they create; 2. capital is deep enough to write billion-dollar cheques; 3. energy and compute are cheap and fast to build; and 4. the rules are light enough to move at the speed of the technology. Measure this strategy against those conditions and the pattern is clear: On founder and employee economics – the single biggest reason talent leaves Canada – it does nothing now. The one capital-gains idea it raises, a reinvestment rollover, is deferred to a study due by Budget 2026. On capital, it makes the government the venture capitalist instead of unleashing private capital to back Canadian companies. On energy, it promises to double the grid by 2050. The build is needed this decade – and the strategy offers no permitting reform to get there. On regulation, it adds a new layer – a trusted-AI certification program, watermarking, plus new privacy and online-safety laws – and compliance always lands hardest on the startups least able to carry it. And it adds a dozen new programs on top of the 130-plus innovation programs founders already can't navigate. The answer was always fewer and faster, not more. Prosperity is not something that the state can spend into existence. Prosperity is what happens when you clear the runway and let the free market work. No government can subsidize its way past the friction it is responsible for creating. Playing to win would look different. It would look like: --> Let founders defer capital gains reinvested in Canadian companies, and fix how we tax employee equity – now, not in a future budget. --> Treat energy and permitting like the emergency they are. Approve power and data centres in months, not years, and build at wartime speed. --> Set a hard speed limit on regulation: apply the laws we have, and clear new products fast. --> Collapse the ~130-plus innovation programs into fewer than ten – and cut the friction founders hit at every step. Canada invented modern AI. We have the talent, one of the cleanest grids in the world, and the research base to win. The opportunity is ours to lose. This strategy is a genuine start – but a country that wants to win doesn't plan to be the world's best customer. It plans to build the companies the world cannot live without.

Carney has unveiled a new council to combat antisemitism in Canada. The lineup includes a former Liberal minister who is a Muslim, a DEI executive who is also Muslim, a progressive lawyer interested in social justice issues who represented pro-Palestinian encampment activists, an LGBT activist, and an Olympic speed skater. Only one member is Jewish: Senator Marc Gold, has a long public record in Jewish communal and pro-Israel advocacy. Seems like expertise in antisemitism was considered optional. If the goal was to reassure Jewish Canadians, this is a bizarre way to go about it. pm.gc.ca/en/news/news-r…

Introducing Impeccable 3.5, the best way to design in production: iterate on real UI with your AI agent, in the codebase you actually ship. Turns out many popular design skills, including Impeccable and Anthropic's frontend-design, weren't actually very good at...design (the workflow was valuable, but the output didn't magically make LLMs like GPT great designers). We measured it across thousands of generations: 74% of pages used the cream AI-default background, 76% reached for extreme letter-spacing, 90%+ failed the contrast floor. So we started fixing slop systematically, specific to each model. The skill now compiles rules for the exact defects each model makes, instead of shipping one generic file to everyone. The biggest jump is in GPT-5.5 and Codex. Also new: ◆ It now knows the difference between a new project and an existing one. Existing codebase, it reads your design system and preserves your identity. Greenfield, it seeds a fresh palette from 129 hand-curated anchors so every cold start doesn't drift to the same safe colors. ◆ Live Mode is now in beta, and works at two scales. Type a direction into the new Steer bar, or speak it, and the agent reads the whole page and edits it in place. Or pick a single element, steer it with a sub-command, live-edit any copy, and accept the variant straight back to source. Insert mode scaffolds brand-new elements between the ones already there. Recovery survives HMR, hidden heroes, and dev-tool overlays. ◆ A rebuilt anti-pattern detector. Torn off jsdom and onto a real CSS cascade resolver: roughly 20x faster, dependency-free, and now small enough to run inline inside the skill, not just the CLI and extension. 14 new rules, 41 total. ◆ The skill keeps itself current, checking once a day and offering to update. Plus /impeccable init and a bare /impeccable that reads your repo and tells you the next move. Free, open source. Claude Code, Codex, Cursor, and more. impeccable.style

A friend just revealed to me that the middle name of Benoit B. Mandelbrot is Benoit B. Mandelbrot.

@_avdept This would be awesome.

Working on my own launcher with tools/features I need in my everyday job. For now added full gh and google calendar support along with basic files and some of omarchy's menus like themes All powered by both - cursor and hotkeys + shortcuts/aliases