@BenjaminDEKR That explains the house cleaning and new employees brought on today.

@justinspratt @MatthewBerman Totally disagree. Camera works way better than you do. Way way way way better, in fact, at night. No need for other sensors. If you say you need other sensors than you must immediately stop driving at night.

they made their "absolutely" theme claude coloured lmao https://t.co/HtZCorzeaB

@RobertSkvarla true. Fast & Furious can’t carry the burden alone

Time to make some mini groyps 💚 https://t.co/e7iMNUzQlg

the coolest thing about kids is the way they have the same exact politics as their parents, absorbing these uncritically throughout their journey into adulthood 💞

@allenholub To summarise: the real mistake was treating a non-deterministic system as if it were deterministic. That is how this technology was introduced, and almost no one challenged the framing. Now every consumer and every company is discovering the difference the hard way.

You can now run Perplexity Computer directly from Slack. → Install from the Slack App Marketplace → Create workflows using context from your channels → Everything syncs to Computer on web automatically No switching tabs. No copy-pasting. Just work where you already work. https://t.co/YT2h3VT9XB

Slack is going to be the interface for AI in the enterprise. And Perplexity Computer fits in very naturally into that interface. The next multi billion dollar and trillion dollar companies are all running on Slack and will see more tasks delegated to AIs than humans.

@aixarizzo Carrot and stick technique

@yudapearl Our world models are action conditioned, and hence causal. The concept of world model for planning goes back to the 1950s in optimal control (before I was born). I didn't just discover it. But training action-conditioned world models from sensory inputs (like video) requires new techniques.

Qodo outperforms Claude Code Review by 19% higher recall and costs 10x less per review. Why this matters for AI devs: @QodoAI just published a comparison using their open benchmark: 100 real pull requests, 580 injected issues, 8 production repositories across TypeScript, Python, JavaScript, C, C#, Rust, and Swift. All three tools hit 79% precision, meaning the quality of individual findings is comparable. The gap comes from recall. Claude Code Review sits at 52%, Qodo Default at 60%, and Qodo Extended at 71%. Great to see attention to recall which matters a lot in this use case. Precision can be tuned post-processing, but recall reflects how deeply a system understands the codebase and reasons about cross-file interactions. If an issue is never flagged, no amount of filtering helps. Even more impressive is how the specialized agent catches the remaining ground truth. Qodo Extended deploys a multi-agent system where specialized agents focus on different issue categories (logical errors, best practices, edge cases, cross-file dependencies), then merge outputs through verification and deduplication. It also leverages model diversity across OpenAI, Anthropic, and Google rather than being locked to a single provider. On cost, Qodo reports significantly lower per-review pricing compared to Claude's token-based model at $15-$25 per review.

Shane should’ve let Lemmy have at least one bite of that sandwich

@allenholub I think you’re arguing against the wrong problem. The AI industry narrative isn’t about non-deterministic technology. It’s about intelligence. An entire generation of researchers, policymakers, and consumers are being sold stories about AGI and superintelligence, often in overtly anthropomorphic terms. Most of them have never even encountered the concept of non-deterministic systems, despite the fact that probabilistic computation and imperfect models have been part of software engineering for decades. So no, non-determinism was never the crisis. The real problem has been speculation, hype, and a remarkable collapse of intellectual honesty about what these systems are and what they can actually do. That has failed the entire conversation.

@nosleepjon - multiple DEX selling wrapped versions (Diesel, Unleaded) - supports multiple wallets (cars, motorbikes, plain)

2023 - write an AI substack (ignorance dot ai) 2023 - attend @aiDotEngineer 2024 - become AI engineer at series B startup 2025 - join @OpenAI to shape API DX 2026 -

2023 - write an AI substack (ignorance dot ai) 2023 - attend @aiDotEngineer 2024 - become AI engineer at series B startup 2025 - join @OpenAI to shape API DX 2026 -

Automations are now GA. You can now: • Set the model and reasoning level • Choose if runs happen in a worktree or existing branch • Reuse workflows with templates Automations are great for recurring tasks — daily repo briefings, issue triage, PR comment follow-up, and more. https://t.co/YlvQauuMqI

Automations are now GA. You can now: • Set the model and reasoning level • Choose if runs happen in a worktree or existing branch • Reuse workflows with templates Automations are great for recurring tasks — daily repo briefings, issue triage, PR comment follow-up, and more. https://t.co/YlvQauuMqI

Pushing the frontier. Come join us if you want to be part of it, my DMs are open.

How many Codexes / Codices / Codii are too many? https://t.co/vH5GbHzWOf

When this wins best editing on Sunday https://t.co/UiwIw7ilNe

The Shield premiered today in 2002! This groundbreaking show was the first from our newly formed MiddKid Productions and starred Michael Chiklis as the corrupt, must-watch Detective Vic Mackey. Watch on Hulu: https://t.co/bzWTjWeAG9 @MichaelChiklis @CCHPounder @cathycryan https://t.co/AK8gEAQyOW

@TIME AI is just software. There’s no super-intelligence. That’s just Silicon Valley VC investors folklore. Don’t fall for it.