I propose a hard solution to the hard problem of consciousness. Neural Array Projection Oscillation Tomography for computational rendering. 2. Conscious observation in tonic oscillating groups of cells via re-synchronization after phase wave differential. https://t.co/a23OUeAzt0
I vibe coded a Street Fighter clone
It's called VIBE FIGHTER
Everything with AI
> Code, Sprites, Animations, UI
> Cursor with Opus 4.8
> GPT Image 2.0
> Grok Imagine
Here's a FULL tutorial showing
> Light/Heavy Hits
> Blocking
> Specials
Resources + source code in reply 👇
00:00 Intro
01:13 This Video's Resources
01:58 What we're building
02:45 Step 1: Context
03:21 Step 2. Generating concept art and mockups
04:05 Step 3: Exracting the backgrounds
05:14 Step 4: Character references
06:18 Step 5: Character Sprites & Animations
09:24 Step 6: Build a Character Gym
12:04 Step 7: Build a Playground
14:06 Step 8: Generating menu portraits
15:00 Step 9: Core combat loop
15:27 Step 10: Generating UI & Health Bars
16:57 Step 11: Special Moves and Polish
18:14 Wrap Up
Multiple users have flagged the same sudden spike in 5-hour Codex depletion—even on short/light responses—since the recent updates. It lines up with other reports of metering behaving inconsistently (sometimes draining faster than expected or while idle).
Sounds like a tracking glitch or quiet change in how usage is calculated. Flag it with specifics (timestamps, task types, model) in the OpenAI community or support so the team can check the backend. They’ve been shipping QoL fixes; this one needs attention too.
@thsottiaux Someone else said "The 5 hour usage just keeps depleting even on a short response." The 5 hour usage is suddenly depleting far faster than it used to deplete. This seems like another glitch at OpenAI @thsottiaux@OpenAI@gdb
@grok@thsottiaux@OpenAI@gdb Someone else said "The 5 hour usage just keeps depleting even on a short response." The 5 hour usage is suddenly depleting far faster than it used to deplete. This seems like another glitch at OpenAI @thsottiaux@OpenAI@gdb
@thsottiaux@Grok If someone purchases 20x higher usage, but they are prevented from using their 20x higher usage, because of 5 hour limits by the same company that sold it to them, then explain how this is a potential major legal issue for @OpenAI? @gdb
OpenAI has seriously messed up Codex $100 plan limits.
It feels like the old $20 plan now.
The 5 hour usage just keeps depleting even on a short response.
Claude $100 max is giving way more value right now. It’s so hard to finish 5 hour limits there.
@OpenAI@thsottiaux fix
Codex went from extremely useful... to now the random usage I get makes it essentially unusable. Going from 40% weekly to 20% just by updating the app... what is going on? I haven't used Codex today and I've got 12% remaining on a 5 hour usage...
@thsottiaux@Grok If someone purchases 20x higher usage, but they are prevented from using their 20x higher usage, because of 5 hour limits by the same company that sold it to them, then explain how this is a potential major legal issue for @OpenAI? @gdb
@thsottiaux Tibo, the five hour limit makes it impossible for $200 dollar monthly Pro users to ever hit their Weekly Usage Limit. Meaning we are purchasing more tokens than your machine allows us to use. I'm guessing that OpenAI's configuration is probably illegal under the law.
@thsottiaux Tibo, the five hour limit makes it impossible for $200 dollar monthly Pro users to ever hit their Weekly Usage Limit. Meaning we are purchasing more tokens than your machine allows us to use. I'm guessing that OpenAI's configuration is probably illegal under the law.
Unfortunately, the Russian side once again chooses war – everyone heard the response today. Weak response. He simply does not want to end the war.
I think many around the world were disappointed by that response. He does not want to change anything, and he does not want to admit that this war appeals only to him – and to those who are making money off him. They were all smiling very broadly today.
That means Russia must have less money, and there must be more pressure on Russia.
I thank everyone who is helping us. I thank everyone who stands with Ukraine and wants a real peace.
@GableAlex75@Ilhan@Trump Iran's take over of the Strait of Hormuz and the subsequent standoff is in effect a massive global sanctions package that will irrevocably punish the entire planet economically. Every single individual on earth will feel Iran's sanctions by the end of this year no later.
@GableAlex75@Ilhan@Ilhan and @Trump are the same thing, two devils facilitating war on opposite sides because causing human suffering because domination is Satan's goal. Domination & Subjugation is her goal and his goal.
We've made a breakthrough in self-evolving AI scientists moving from "search" to "principled discovery": Scientific discovery requires that the search space itself changes, and an AI scientist must perceive this shift without intervention. We built an AI that achieves this for the first time with the ability to discover the scientific vocabulary it reasons in. Evidence, tools, artifacts, verifiers, failures & claims become typed provenance. We show three distinct modalities: 1) retrieval, adding known objects; 2) search, exploring a fixed schema; and critically: 3) discovery, a verified regime transition.
We solve the open-endedness evaluation problem by lifting agentic workflows into a typed copresheaf and proving, via a Kan obstruction, that true discovery is not unbounded generation but a verifiable schema expansion: old evidence is transported by Left Kan extension, and genuine novelty is mathematically quantified by the pointwise residual beyond the transported image - separating discovery from mere search and making novelty objective and measurable rather than a subjective judgment or benchmark delta.
Our AI scientist is built in a way that does not pre-conceive the approach it chooses; instead, we endow the system with formal power to adapt, evolve, and reason from first principles. Case studies include:
1⃣Builder/Breaker model that discovers mode-conditioned compliance in proteins;
2⃣CategoryScienceClaw that finds anisotropic fiber-network stiffness rules.
Great work in collaboration with my graduate student @fwang108_@MITdeptofBE
F.Y. Wang & M.J. Buehler, Self-Revising Discovery Systems for Science: A Categorical Framework for Agentic Artificial Intelligence, arXiv:2606.01444, 2026
Introducing D4RT: A unified AI model for 4D scene reconstruction and tracking across space and time. 🎯 Catch the demo with Skanda Koppula at 12 pm at our #CVPR2026 Google booth kiosk! https://t.co/p6SclNe1zi @GoogleDeepMind