Kristian T

Brother

@0xSero Quants available for us 5090 plebs?

Really good model, enjoyed using it. Feels like a complete package.

Claude still has the best /loop mechanism by far, its crazy what you can achieve with a well refined objective and a locally scheduled loop

Was using WISRP Flow a tonne over the past handful of months, had awesome productivity from it, but found myself blabbering on too much to the agents. Typing is more refined, my brain seems to do an extra filter process before typing, whereas brain to speech has little filter

@StopBrokeNow @unusual_whales It's confirmed for the chip workers, over 25B total.

@unusual_whales $340K average bonus per worker would be ~$24B total, bigger than Samsung Electronics' annual op profit. The actual deal is closer to a few thousand dollars. Whatever number got tweeted has too many zeros.

Samsung's largest union voted in favor of a compensation deal that will hand chip workers an average bonus of about $340,000

@calvinfroedge You misread how the markets work. Markets at the macro level go up and down on changes to liquidity. The thing with geopolitical issues / bswans is they can be worked around with liquidity injection. Don’t think about the headline - think “will liquidity increase or decrease?”

- Bet on Strait of Hormuz closure, a black swan event they said could never happen because it would be so damaging to the global economy - Be correct - Make no money - Markets have sharpest rally in financial history

@thsottiaux would love a better UX in Codex(mac) when working with automations with high frequency runs. Digging into each of the "Previous runs" one by one is a bit painful. Would also love memory/smarts around how I create automations - that modal seems to always reset settings

@brettmacro The consumer is what was holding up the economy in the past, but not anymore - now economy is perked up by excessive government spending, and corporate capex. Consumer can be hurting, but currently that doesn’t matter. For now…

When someone says "The stock market doesn't equal the economy", this is what they mean. Consumer Sentiment just hit the lowest point on record (70 years of data).

@gabrielchua Would have loved to be there!

Good morning, Sydney - Codex

@jonathan_wilke @cursor_ai Agreed. I currently have subscriptions / access to just about everything, including a beastly local-AI setup. Cursor is what my hands automatically reach for - but the whole ecosystem actually with their cloud agents and bugbot is really solid

I'm gonna say it: As it stands today, @cursor_ai is the best agentic coding tool out there. No jumping between terminal and editor, nice UI and you don't have to switch the tool if there's a new model. change my mind.

@gabrielchua Would have loved to come along, this week didn't work out for me though. Hopefully next time!

Hello Sydney 👋 🇦🇺 GPT-5.5 is here, and the OpenAI Codex & Startup teams are bringing the vibes, tokens, and builder energy to town next week. Catch us here: > OpenAI x January Capital x Lyra x Relevance AI Hackathon [luma.com/aq1yr5vc] > Builder session with Sydney Computing Society - SYNCS/ University of Sydney > Vercel x OpenAI Builder Day [lnkd.in/gU5FH_nX] > Coffee, Coworking and OpenAI Codex with Build Club [luma.com/1hi8t8kw] > OpenAI Codex Hackathon - Sydney, co-hosted with University of Technology Sydney, UTS Startups and Arafat Tehsin, Codex Ambassador for Sydney [luma.com/or8icykr] > Going Global: AI Founders Lunch with OpenAI, Crane Venture Partners and Liminal [luma.com/i5j5egfp] > January Capital x Accel x OpenAI x Airwallex Pitch Day [luma.com/hdw4cp7j] I’ll also be in Melbourne on 30 April for our Codex Community Meetup, hosted by Dr Sam D., Codex Ambassador for Melbourne, and the crew at MLAI: luma.com/yhc5wr8h I also heard Thomas Jeng will be recording an episode of the Startup 360 (Startup Daily) podcast with Simon Thomsen Most of all, I’m excited to finally meet my wonderful teammates in Australia IRL 💙 PS: Some of the events are RSVP-only or registrations have closed, but happy to chat if you’re building something really cool with Codex, GPT-5.5, or GPT Image 2.

@chribjel People need to remember the harness is also super important, esp for coding. Claude Code / Codex / Cursor are doing A LOT for the agent orchestration and overall result quality and experience. When jumping to the local AI ecosystem need an equally solid harness setup

gives «benchmaxxing» vibes How is it actually to use it??

@neural_avb Just hire a cloud GPU for a handful of hours (from an aggregator like vast AI) - will cost you $10-$20 to get a general feel on how good it can get, before forking out to invest hard in your local setup

Qwen3.6-27B is the first model that made me want that Mac Studio My current mac runs the 4bit version at just 15 tok/s. Loading the initial opencode system prompt takes like 30-50 seconds. Is there a draft model I could be using inside LM studio for faster inference?

@nlevnaut @LottoLabs In real world engineering cases you’re dealing with a skilled practitioner giving models clear bite sized tasks with guardrails. we are very much at the stage of “good enough” for those type of workflows.

@LottoLabs the wild thing is the frontier models are kinda starting to feel "good enough". seems like if we get the good ending, the frontier will be commoditized

Assume qwen 3.6 27b isn’t actually opus level or even sonnet level knowledge It’s much better than sonnet 3.5 level And that was sota not long ago, a loved model We’re in crazy times

@dexhorthy Thanks @dexhorthy - love the things that you're doing for the AI engineering community

Harness Engineering Without the Hype: 🦄 AI That Works #54 x.com/i/broadcasts/1…

@mattpocockuk Simply saying "make this code more 1337" about a specific block of code often has surprisingly good results

What 'advanced' AI coding techniques are you using? I.e. what do you feel like you've discovered that no-one else knows about yet?

@beruu99_ @forallcurious “We are the first” is one of the most plausible answers to the Fermi Paradox

@forallcurious Say our galaxy, not earth. We are 4 ly away from stars in our galaxy, and we're 2.5 million ly away from Andromeda. We're not especially alone in a huge void, we're simply in a region of space with ~half the average cosmic density. Mby one reason we could be early universe life

🚨: New evidence shows that the Earth is trapped in a void 2 billion light-years in diameter

@sudoingX Qwopus 27b v3

best model on 24gb vram in april 2026?

@0o0Oo0O0oOO @cryptodude999 Their jerseys now are shocking quality, I only buy retro kits these days. The new stuff falls apart after one gentle cycle in the washing

@cryptodude999 They need to stop selling synthetic shit and go back to cotton. That's all.

$NIKE Go woke, go broke.

@outsource_ Exactly what I was hoping for - going to give it a try on my 5090 🎉

🚀 NEW GEMMA 4 31B TURBO DROPPED Runs on a SINGLE RTX 5090: ⚡️18.5 GB VRAM only (68% smaller) 🧠51 tok/s single decode 💻1,244 tok/s batched 🤖15,359 tok/s prefill ← yes, fifteen thousand 🚨2.5× faster than base model with basically zero quality loss. It hits Sonnet-4.5 level on hard classification tasks… at 1/600th the cost. Local models are shipping faster than we can test 👇🏻 🔥 HF: huggingface.co/LilaRest/gemma…