Ian Hansel retweeted
Gaming industry is dead.
Remember 2 months ago when no one believed me? Now we have Vibecoded decompiled Frankenstein halo call of elder mine scrolls portal craft remix palooza.
The economics and business model behind every game studio is GONE.
OpenAI is going to build Apple gamecenter on top of ChatGPT sites, and in building the social layer, will then sweep the entire industry and we will never see a major game release built by humans again.
WOW:forever and GTA 6 will be the last major games to ever be published.
Vibe gaming is here.
Final form will resemble what we are beginning to see now, where people share & monetize games that are closer to memes and “content”, not full AAA classical game experiences.
Books -> Tweets
Movies -> YouTube -> TikTok
Games -> Viral Vibegamed MMO memes
If you’re reading this, you have likely seen dozens of these new vibe games already, and very likely played or even made your own.
You’re already playing next generation
I’m pretty sure gaming as a category is dead because everyone is just gonna build their own games
Games don’t have ongoing maintenance cost like companies replacing SaaS vendors
Perhaps an argument for wanting someone else to build an experience for you, easier to watch a movie than create one.
But video games and vibe coding are so close the difference between creating the form and playing in a form is getting closer and closer to zero.
So why play someone else’s game?
Just pay chatgpt $20 and make you own
Ian Hansel retweeted
In 1956, Cybernetics was renamed "AI" by logic programming experts. Now Trump wants to rename AI again. Let's revive the old name! Modern AI is about neural nets & deep learning, closer to Cybernetics than to 1956 AI. Plus, Cybernetics sounds better than Artificial Intelligence.
Ian Hansel retweeted
I've come up with a new way of exploring large datasets with UMAP called Latent Craft.
Fly through this amazing dataset and mine for images you want to collect.
all 1 million explorable in the browser! huggingface.co/spaces/enjalo…
Uploaded a dataset of 1,080,814 public domain images, mostly from 19th-century books, to the Hub.
huggingface.co/datasets/bigl…
Restkontingentscham - meaning “the shame of letting your Codex allowance reset before you’ve used it all.”
3
263
Ian Hansel retweeted
Say your AI agent is about to negotiate with your supplier's AI agent.
You tested yours. They tested theirs. Nobody tested what happens when they meet...
🧵
Ian Hansel retweeted
Introducing Prime Agent:
A self-improving RLM harness for coding and long-running autonomous tasks.
Designed to be both token-efficient and expressive through programmatic tool calling, context as a variable, multi-agent messaging, and a self-modifiable harness state.
Ian Hansel retweeted
Its crazy how the consensus shifted away from Claude being the favorite to Codex
It’s not just Vibes. Things favoring codex:
- Better harness (Codex Desktop)
- Better pricing: more things included within subscription (fast mode etc). bonus - constant resets from Tibo 😅
- Less refusals / false positive safety triggers
- You can use subscription freely wherever you want (Opencode, Amp, etc)
When you add all these up it makes codex significantly more attractive. There is likely an industry lag of people still using Claude but feels like it’s gonna swing back
Ian Hansel retweeted
turned my obsidian notes into fish 🐟🐠
draw a little stream and the school swims it like a timeline: newer notes at the mouth, older ones drifting toward the glow. hover a fish and it blows a bubble with the note tucked inside.
available for my subscribers <3
Ian Hansel retweeted
New uses require new interfaces
These are the three things I have been trying out to control Codex. Given how useful voice mode is, the Teenage Engineering Ting walky-talky has been a surprising favorite. The Codex micro is beautiful & fun, but too little info vs the Stream Deck
Ian Hansel retweeted
turn your xbox/PS controller into your Codex controller.
Ian absolutely cooked with this one.
control-deck.com/
get around it @thsottiaux @sama @embirico @edbayes
Ian Hansel retweeted
Replying to @JensenHuang @nvidia
Ian Hansel retweeted
You need to be datavizmaxxing
If you aren’t using your on-tap 300 iq robominions to actually understand what you are building, you are losing anon
Ian Hansel retweeted
My 2hr workshop on Open vs Closed models, reward hacking, benchmaxxing & RL is out!
1. Closed vs open models
2. Throughput maxxing but accuracy minimizing
3. Benchmaxxing & cheating
4. Distillation & RL
5. Stopping reward hacking
6. @UnslothAI Dynamic Quants
Details:
1. If reasoning wasn't discovered via o1-preview - would all AI progress grind to a halt via a S shape? Reasoning made doubling times now 3.5 months instead of 7 - so just wait 3.5 months for the next best model! Ways folks might regulate open source - like a driver's license for AI
2. Inference providers maximize throughput and speed, but accuracy degrades - OpenRouter publishes stats on accuracy, and the largest gaps can be 20% or more!
3. METR, WeirdML, Deep-SWE, FrontierCode, SWE Bench Pro - what are good & bad benchmarks? False positive / false negative rates - how about daily benchmarking since Codex & Claude have perf regressions. How can we regressions to predict next model releases as well?
4. Open model labs use distillation partially, but need RL to create the reasoning traces to complete the process - hard vs soft distillation and how to automate RL.
5. Common examples for reward hacking + ways to stop it - internet filtering, classification systems, timing / editing global variables, real world examples + more!
6. Why software matters more than hardware - the limits of FP4 and GPUs vs ASICs and torch.compile vs kernels and megakernels and more, and why quantization and memory optimizations are important
youtube.com/watch?v=uIiA6Dqu…
Ian Hansel retweeted
happy official launch day 1! launch week opens with updates to marimo-pair, now in VS Code. pair opens up a new medium for agent-native data and computational work: you + your agent, collaborating over one interactive, flexible canvas.
more below ⬇️
marimo.io/pair
Ian Hansel retweeted
Concentration of power, capabilities and economic wealth is the biggest risk in AI. We need open science and open-source more than ever!
Ian Hansel retweeted
We collaborated with our friends at Marimo to bring @UnslothAI Studio to Molabs with online GPUs!
Powered by 96GB NVIDIA RTX 6000 PROs via Coreweave's Cloud, large finetuning & inference runs can be run in 12 hour sessions! Notebooks at github.com/unslothai/noteboo…
Announcing a public preview of molab, our cloud-hosted marimo notebook workspace: now with GPUs! 🧵
molab.marimo.io/notebooks
Made this Windows95 (and Palm) throwback with @lennysan data to get LennyOS - might have to go back to WinAmp for listening to podcasts now. Try it in link below
1
87




























