🚨 openai just dropped 20 plus things at devday
- dots: alwayd on agents
- gpt-6.1 sol: near astra speed at about one fifth the price
- ultrafast: 300 tok per sec in codex, 6 times faster in api
- pro $500: more usage plus ultrafast
- private intelligence: keeps your data safer
- codex cloud, voice cli, desktop code review
- codex security cloud: scans repos and helps fix them
- decisions api (luna): limited preview
- agents api with computer use, bedrock on aws
- better plugins, sites plugins, mcp events
- chatgpt space, pages, teams, slides soon
- chatgpt in slack and teams, meetings plugin
- sign in with chatgpt, openai marketplace
not all free or live for everyone yet
- dots: alwayd on agents
- gpt-6.1 sol: near astra speed at about one fifth the price
- ultrafast: 300 tok per sec in codex, 6 times faster in api
- pro $500: more usage plus ultrafast
- private intelligence: keeps your data safer
- codex cloud, voice cli, desktop code review
- codex security cloud: scans repos and helps fix them
- decisions api (luna): limited preview
- agents api with computer use, bedrock on aws
- better plugins, sites plugins, mcp events
- chatgpt space, pages, teams, slides soon
- chatgpt in slack and teams, meetings plugin
- sign in with chatgpt, openai marketplace
not all free or live for everyone yet
👾5
Open Ai
🚨 openai just dropped 20 plus things at devday - dots: alwayd on agents - gpt-6.1 sol: near astra speed at about one fifth the price - ultrafast: 300 tok per sec in codex, 6 times faster in api - pro $500: more usage plus ultrafast - private intelligence:…
Check Comments For More Details
👾39
JUST IN: GPT-6 Astra cracks unsolved 217-year-old Napoleonic military cipher, revealing a secret briefing sent to one of Napoleon’s generals ahead of Austria’s 1809 invasion.
👾38🔥1
JUST IN: SpaceXAI launches the Grok Bot Marketplace, allowing users to add specialized AI agents for engineering, research, sales, marketing, and more.
👾5
BREAKING 🔥: Google announced Gemini 4 Argon, a new frontier model for "complex workflows across real-world software engineering, enterprise knowledge work like legal and finance, and cyber defense."
77.9% score on DeepSWE v1.1 is a new SOTA. It performs better then GPT-6 Astra, Opus 5.5 and Fable 5.1 across many benchmarks.
Rolling out soon starting with paid API customers and Google AI Ultra subscribers.
Soon! 👀
77.9% score on DeepSWE v1.1 is a new SOTA. It performs better then GPT-6 Astra, Opus 5.5 and Fable 5.1 across many benchmarks.
Rolling out soon starting with paid API customers and Google AI Ultra subscribers.
Soon! 👀
👾4
This media is not supported in your browser
VIEW IN TELEGRAM
SPACEXAI 🔥: Grok Bot becomes proactive and can now suggest helpful things on its own.
Users will be asked to configure a new Primary Bot or select an existing Bot as Primary (ex., your Chief of Stuff).
The Primary Bot feature is rolling out gradually now!
Users will be asked to configure a new Primary Bot or select an existing Bot as Primary (ex., your Chief of Stuff).
The Primary Bot feature is rolling out gradually now!
👾38
Aleph Alpha released Kolibri, a 78B-parameter, 3.46B active, MoE open-weight model with 1M context window, distributed under the Apache 2.0 license.
Aleph Alpha is a German AI lab, and it is probably the first general-purpose model from the EU, apart from Mistral, to achieve this level of performance!
> "The model supports an explicit reasoning mode and tool calling. It is optimized for long-context and inference efficiency."
> "The Aleph Alpha team developed a bilingual German/English tokenizer and focused on including organic German data throughout the training process of the model, so that 21.3% of the pre-training tokens are German."
> "Kolibri was built with the EU AI Act, the General-Purpose AI Code of Practice and the GDPR in mind from the ground up, with copyright law being a focus of our work to trustworthy technology."
More EU labs joining the party 👀
Aleph Alpha is a German AI lab, and it is probably the first general-purpose model from the EU, apart from Mistral, to achieve this level of performance!
> "The model supports an explicit reasoning mode and tool calling. It is optimized for long-context and inference efficiency."
> "The Aleph Alpha team developed a bilingual German/English tokenizer and focused on including organic German data throughout the training process of the model, so that 21.3% of the pre-training tokens are German."
> "Kolibri was built with the EU AI Act, the General-Purpose AI Code of Practice and the GDPR in mind from the ground up, with copyright law being a focus of our work to trustworthy technology."
More EU labs joining the party 👀
👾38👍2👎1
This media is not supported in your browser
VIEW IN TELEGRAM
🚨 BREAKING: AI can now take apart almost any game or software and rebuild it. People are using it to mash games together.
- Minecraft mechanics inside GTA.
- Spider-Man inside a Batman game.
- Mirror's Edge parkour inside Skyrim.
It all works, and it only takes prompting.
How it started 👇
> OpenAI released GPT6 Astra.
> On a benchmark for reverse engineering software from its binary, it succeeds over 99% of the time within four attempts.
> Claude Opus 5.5 reportedly does even better.
It has already decompiled
- Mario Kart 64
- Halo
- Kingdom Hearts and more.
The same logic applies to Photoshop, Office and operating systems.
- Minecraft mechanics inside GTA.
- Spider-Man inside a Batman game.
- Mirror's Edge parkour inside Skyrim.
It all works, and it only takes prompting.
How it started 👇
> OpenAI released GPT6 Astra.
> On a benchmark for reverse engineering software from its binary, it succeeds over 99% of the time within four attempts.
> Claude Opus 5.5 reportedly does even better.
It has already decompiled
- Mario Kart 64
- Halo
- Kingdom Hearts and more.
The same logic applies to Photoshop, Office and operating systems.
👾37
This media is not supported in your browser
VIEW IN TELEGRAM
Three ways people use it:
1️⃣ PASS-THROUGH MOD
The AI builds a bridge between two running games.
> A Minecraft player places TNT
> the bridge tells GTA to spawn an explosion at the same spot.
> Both engines stay in control, so collisions and physics can break.
2️⃣ REBUILDING THE WHOLE GAME
The AI works out the
- animations
- maps
- models
- mechanics
then rewrites the game in a new language, usually Rust.
You must own the game
but once it runs on your own engine you can change anything.
Results so far:
- World of Warcraft on a skateboard
- Call of Duty merged with Minecraft
- Super Mario Galaxy in VR
- PS5 games on PC.
3️⃣ PORTING ONE MECHANIC
The AI lifts a single system from one game and drops it into another.
> Mirror's Edge parkour went into Skyrim in about a day.
> Skate 3 physics went into Call of Duty.
> It is copy and paste for game design.
1️⃣ PASS-THROUGH MOD
The AI builds a bridge between two running games.
> A Minecraft player places TNT
> the bridge tells GTA to spawn an explosion at the same spot.
> Both engines stay in control, so collisions and physics can break.
2️⃣ REBUILDING THE WHOLE GAME
The AI works out the
- animations
- maps
- models
- mechanics
then rewrites the game in a new language, usually Rust.
You must own the game
but once it runs on your own engine you can change anything.
Results so far:
- World of Warcraft on a skateboard
- Call of Duty merged with Minecraft
- Super Mario Galaxy in VR
- PS5 games on PC.
3️⃣ PORTING ONE MECHANIC
The AI lifts a single system from one game and drops it into another.
> Mirror's Edge parkour went into Skyrim in about a day.
> Skate 3 physics went into Call of Duty.
> It is copy and paste for game design.
👾36
playgta5.com web browser gta5
moon-zombies.pages.dev BO1 moon zombies in web browser
surfd.net cs surfing on the web browser
global-terror.net Skate Rust
moon-zombies.pages.dev BO1 moon zombies in web browser
surfd.net cs surfing on the web browser
global-terror.net Skate Rust
👾37
OpenAI published a repository with novel solutions to a range of math problems solved by their next frontier model.
The vast majority of results were obtained with the same procedure using an unreleased internal OpenAI model.
On average, each result used three hours of ChatGPT Pro thinking compute with that model.
Over the course of the evaluation, the model was posed approximately 4,000 problems.
This media is not supported in your browser
VIEW IN TELEGRAM
opus 5.5 is insane as a motion designer
this 15-second video turns one orange dot into an entire motion system
toggle → typography → spring graph → shape morph → kinetic type → halftone → end card
here's the prompt to recreate the idea ↓
https://t.me/CorePrompts/164
this 15-second video turns one orange dot into an entire motion system
toggle → typography → spring graph → shape morph → kinetic type → halftone → end card
here's the prompt to recreate the idea ↓
https://t.me/CorePrompts/164
👍1
BREAKING 🔥: Anthropic released Claude Haiku 5.5, its fastest and cheapest model to date.
Haiku 5.5 costs 75% less to run than Haiku 4.5!
It is the first Haiku model to have an adjustable effort setting.
On benchmarks, it is a big leap in performance at Computer Use and Agentic Coding across the group.
👾36