Michael joy

💻 AI PM | Building in Public
💰 Currently building:
• AI Employees
• AI SaaS
• Enterprise AI Solutions
🚀 Sharing the latest on AI trends, tool

[Michael joy @Michaeljoy1213]

GPT 6 Sol It is a rubbish, often modifying previously determined tasks, which is very far behind Claude 5.5

[Michael joy @Michaeljoy1213]

@thsottiaux Refund the money, I don't want to play at your house anymore, it's disgusting.

[Michael joy @Michaeljoy1213]

All the good open-source models, I help.

[Xiaomi MiMo @XiaomiMiMo]

Introducing Xiaomi MiMo-V2.6 — Pro & Flash.

Frontier intelligence, all the modalities, built in public.

🔹 Two omnimodal models, advancing through scaled reinforcement learning
🔹 Pro performs on par with Claude Opus 5 and GPT-5.6 Sol across most agent benchmarks
🔹 Pro scores 46 on the Artificial Analysis Intelligence Index — the highest among open-source models
🔹 Stronger coding, computer use, 3D reasoning and creative capabilities
🔹 Open model weights, technical report, RL environments and training code

Blog：https://t.co/JGdwC9Ymvj

[Michael joy @Michaeljoy1213]

@thsottiaux @stemonteduro Now the whole consumption is too fast, don't you admit that you have done something? Anyway, I feel very disgusting.

[Michael joy @Michaeljoy1213]

@thsottiaux @stemonteduro Saying this is like farting, don't you use your brain? Why should users buy it? We are all Pro users and spent so much money. The quota is exhausted very quickly.

[Michael joy @Michaeljoy1213]

GPT 6 Astra feels that its IQ is getting lower and lower, and it cannot complete a task. The quota is consumed too fast. A weekly quota has to wait for 7 days. I am a Pro subscriber, but the experience is really terrible. The consumption is too fast, and the IQ is still very low.

[Michael joy @Michaeljoy1213]

Listening to you now is a bit like farting. The reset mentioned last Tuesday has not been implemented yet. You are complete liars!

[Tibo @thsottiaux]

We were working on the keynote today with @romainhuet and @sama and most of the fun was trying to figure out how to explain it all to you because there is so much good stuff in there that it's a bit ridiculous all in quick succession.

We'll have some things next week already to not keep you waiting so long, but very excited to show you all new things we've been working on and how it will all come together in the coming months.

[Michael joy @Michaeljoy1213]

@sama I now feel more and more that Astra is completely different from what was bragged at the previous press conference. Its IQ is getting lower and lower, and the quota consumption is very fast. Everything is a lie.

[Michael joy @Michaeljoy1213]

@thsottiaux Do you know that now Astra can't complete the task effectively at all, the consumption is too fast

[Michael joy @Michaeljoy1213]

@OpenAIDevs To be honest, the consumption is so fast that there is no way to complete a task quickly

[Michael joy @Michaeljoy1213]

Use Astra + blender + unity to realize the real physical world, and discuss how you can restore the real environment together

[Michael joy @Michaeljoy1213]

@thsottiaux 工作模式和聊天模式以及codex和ChatGPT，这些区分太复杂了，直接一个模式即可

[Michael joy @Michaeljoy1213]

@Frank_web33 Wow, that's amazing.

[Michael joy @Michaeljoy1213]

@HealthRanger Strongly agree

[Michael joy @Michaeljoy1213]

What story are you making up, a lie made up to make a very small number of you use a stronger model for world domination? We don't want to believe the lies of these capitalists, they only make money for profit and all of us!

[Anthropic @AnthropicAI]

We're publishing our most detailed threat intelligence report to date.

It covers how people tried to misuse Claude—for cyberattacks, influence operations, surveillance, biology, and building weapons—and how we found and stopped them.

We disrupted every operation in the report, and used the lessons from them to strengthen our safeguards. Where appropriate, we also shared what we found with authorities and other AI companies.

These cases are not typical: we’re highlighting some of the most sophisticated misuse we’ve seen. But they’re especially important to discuss, because they show us where AI misuse is headed, where our safeguards work, and where they need to improve.

We’re publishing this report so others can spot the same activity on their own platforms, and so we can give the public a clearer view of how emerging threats develop.

Read the report: https://t.co/0EJUnYEgfz

[Michael joy @Michaeljoy1213]

You don't think your current model is so powerful, do you? Your raw model says that multiple edits still won't change other elements. On the contrary, Astra is not as smart as you boast. Except for the faster consumption quota, it can't get the job done at all

[Sam Altman @sama]

I agree with Dario that we need to pace the frontier. This has been a primary topic of discussions we've had at OpenAI in recent weeks.

Committing to having independent evaluators with employee-like access is a great idea, and we will do the same. We'll have more to share soon.

[Michael joy @Michaeljoy1213]

@thsottiaux I called the deep map model in Codex. The deep map model in it should be Imagen 2.5, right?

[Michael joy @Michaeljoy1213]

@thsottiaux I don't think Shadow Plum 2.5 is as powerful as you said. I let it generate the picture for me today, and then every time it edits, it changes other places as well.

[Michael joy @Michaeljoy1213]

@buildwithrajath Hahaha, that's true.

[Michael joy @Michaeljoy1213]

@thsottiaux Has Astra lowered his IQ recently? I asked him to complete a modeling task, but it hasn't been done in a day, and the quality is very low, so it doesn't work at all.