Steve · AI

Building @PrivOSAI | Enterprise AI suite where teams & agents collaborate | | Agentic Hotel OS | Sharing real AI insights from building

[Steve · AI @stev_builds]

Many corporate leaders waste significant time coordinating daily operations across multiple disconnected software tools. This reliance on fragmented communication channels delays approvals and makes it extremely difficult to track real-time project progress.

Transitioning to a unified workspace allows teams to automate repetitive approval workflows securely inside a self-hosted network. For instance, consider an operations team that approves pricing policies instantly within the chat interface rather than waiting for email threads.

Read our complete analysis to learn how consolidating your software stack accelerates business performance.

[PrivOS @PrivOSAI]

https://t.co/iKlcF0Bv8O

[Steve · AI @stev_builds]

There's a real difference between using AI and owning AI

Using AI: your people prompt ChatGPT for emails, summaries, quick analysis. Fine. Useful.

Owning AI: your org knows which processes to hand off, who's accountable, how far an agent can act, and how humans stay in the loop to catch mistakes.

But AI can't do what isn't written down. If the process lives in someone's head, the agent has nothing to run on. It guesses. It drifts. You get chaos at scale instead of leverage.

AI-ready starts with documented workflows. Clear decision rights. Actual governance.

So are you using AI, or does it own how AI operates within it? 🧠

#AI

[Steve · AI @stev_builds]

Someone accidentally exposed Fable 5's unfiltered "inner voice", and apparently it spends its reasoning process muttering to itself.

A user gave Fable 5 an extremely difficult competitive programming problem. Instead of only returning the polished answer, the web interface briefly revealed its hidden reasoning trace.

What showed up was bizarre:

• Repeating phrases like: "DATA DATA DATA. GO."
• Growling things like: "GRRR" and "GAAAH" whenever it hit a difficult section.
• A relieved "PHEW" once it finally found a solution.
• The whole thing looked less like English and more like frantic shorthand from a stressed-out caveman.

The clear, well-structured answers we usually see appear to be just the polished output layer. Behind the scenes, the model seems to "talk to itself" in a compressed internal dialect optimized for reasoning

Looks like it has developed its own private language for thinking... Doesn't this sound scary?

[Steve · AI @stev_builds]

@moo9000 Or let's call it a "bribe"?

[Steve · AI @stev_builds]

@LLMJunky @alexocheema @davidweiss I get what you want to say, I had my self a M3 Ultra cluster but I know also what its deliver. My only point in my question was, what he wanna run with 6 M3 Ultras which would make sense

[Steve · AI @stev_builds]

@LLMJunky @alexocheema @davidweiss A M3 Ultra with 512gigs costs >25k.
Good luck finding 6 even. Beside what you wanna do with a 3tb model and 80gb cable connection, generate 1t/s?

[Steve · AI @stev_builds]

@DeItaone RIP for them who relies on disc's as they have poor internet connection. In my summer house I have 20mb internet and even updates took ages

[Steve · AI @stev_builds]

@0x99Gohan @kylobtc Waiting for an airdrop

[Steve · AI @stev_builds]

@alexocheema @davidweiss Why the hack you would want cluster 6 M3 Ultras?
You can combine 5 devices together, enjoy multiple bottlenecks. If you want to spend that kinda money there are way better solution for inference

[Steve · AI @stev_builds]

@NEARProtocol Not usable to be honest, getting time outs and unusable through put.

If you can solve this, it would be great. Otherwise you can't scale with your customers

[PrivOS @PrivOSAI]

When decisions happen inside a flat message stream, the context that explains them disappears within hours.

PrivOS Chat threads keep each discussion anchored, complete, and readable long after the conversation ends, which matters even more when AI agents need accurate context to act on what your team decided. 💬

How does your team currently preserve the reasoning behind decisions that start as chat messages?

#PrivOS #EnterpriseChat #AIWorkspace

[Steve · AI @stev_builds]

@support_huihui Waiting patiently

[Steve · AI @stev_builds]

@jun_song Honestly, how you want to run DS v4 Flash with a RTX 3090?

[Steve · AI @stev_builds]

Meta just introduced Brain2Qwerty 🤯

A model that turns brain activity into text.

Today we type.
Tomorrow we might just think.
Emails written.
Code completed.
Ideas captured instantly.

From keyboards to brain-computer interfaces.
The future of computing may simply be:
Think → Create → Done. 🧠⚡
AI + Science.

#AI #Meta #Brain2Qwerty

[Steve · AI @stev_builds]

GLM-5.2 is looking pretty weak on ARC-AGI 👀

ARC-AGI — especially the v2 version — has a reputation for exposing models that are heavily “benchmaxxed” and optimized just to look good on leaderboards.

From the looks of it, GLM-5.2 seems to be struggling to keep up with Opus, and it doesn’t appear to reach even GPT-5.5 low reasoning levels.

It’s also not significantly cheaper, with pricing sitting in roughly the same range. 😀

At this point, it doesn’t look like a very compelling option.

[Steve · AI @stev_builds]

@ZixuanLi_ Its optimized for Blackwell Datacenter grade GPUs, so as a Inference provider you can serve much more user with better throughput. NVFP4 quantity are quality wise similar to Unsloth ones

[Steve · AI @stev_builds]

@TamagoFluffy @Cointelegraph Alibaba is listed on NYSE

[Steve · AI @stev_builds]

@Cointelegraph Official Qwopus coming

[Steve · AI @stev_builds]

Claude just introduced Claude Tag to bring AI directly into Slack conversations and assign tasks like working with a teammate.

This is exactly the future we have been building toward at PrivOS.

The next evolution of work is about humans and AI working together inside the same environment, sharing context, understanding workflows, and completing tasks together.

That is why PrivOS was designed as an AI Operating System for Enterprise, where teams and AI agents collaborate in a secure workspace.

[PrivOS @PrivOSAI]

🚀 AI adoption is no longer only for large enterprises.

For many small businesses, the challenge is not recognizing the value of AI. The real challenge is knowing where to start, which workflows to automate, and how to integrate AI without increasing complexity.

🤖 Business AI Scale provides a practical roadmap to help organizations understand how to build AI-ready operations step by step, from identifying opportunities to implementing AI into daily workflows.

Whether you are exploring AI for the first time or looking to scale existing adoption, the right strategy can help your business move faster with confidence.

📂 Discover the Business AI Scale guide and find the roadmap to integrate AI into your organization. https://t.co/jShhFc92lF

#EnterpriseAI #AIAdoption #BusinessAutomation #AIReady #PrivOS