Seb Montoya

Crypto, Stocks, Surf, Books & Rock n’ Roll @RealVision Member 🚀. @SuiCommunity advocated 🚀

[Seb Montoya @Chevy42000]

@aisatyam @NousResearch Yes Sir!

[Seb Montoya @Chevy42000]

Building, Learning the stack @NousResearch Included, and sharing in public. If that’s you too, let’s Connect 🤙🏼

[Seb Montoya @Chevy42000]

@rileybrown @agentnative_ this is great set up

[Seb Montoya @Chevy42000]

The more I learn about Agents, the more curious I get about Crypto’s role. That’s part of what draws me to @SuiCommunity Still learning, but this a future I’m excited to explore.

[Seb Montoya @Chevy42000]

@SubhadipShips Yes go!!🚀

[Seb Montoya @Chevy42000]

#NewProfilePic. @SuiCommunity @SuiNetwork is the way ! @EmanAbio 👀

[Seb Montoya @Chevy42000]

@EmanAbio LFG!!

[Seb Montoya @Chevy42000]

Right now I’m building a system for a Cafe that connects to their existing CRM, tracks inventory, and shows profit per product tied directly into @Square. Using Claude Code and Codex to move through the plan faster.

What’s wild is these tools aren’t just speeding the work up. They’re opening a whole new phase of growth for someone who never came from a technical background. Still early, still iterating, but it’s real progress.

If you’re building something similar, or learning and sharing in public, let’s connect 🤝

[Seb Montoya @Chevy42000]

@Da7_Tech YesSr!

[Seb Montoya @Chevy42000]

@danellisona 💩 probably

[Seb Montoya @Chevy42000]

@danellisona X is definitely a very powerful tool to connect and build community.

[Seb Montoya @Chevy42000]

@erden_xi Cow Farm bro for sure in Colombia 😇

[Seb Montoya @Chevy42000]

@HermesWatcher Love that!

[Seb Montoya @Chevy42000]

100 followers😅📈. The post yesterday about learning Claude Code and Codex got a few replies. What’s one tip you’d give someone just starting?

[Seb Montoya @Chevy42000]

@jah_3li @NousResearch yes Sr!!

[Seb Montoya @Chevy42000]

@Yaorids G’day!

[Joseph N. Aburu @Josephyala]

The next big leap in AI may not come from a smarter model. It may come from everything we put around it.

NVIDIA CEO Jensen Huang describes the “agentic harness” as an exoskeleton around an LLM. It gives the model retrieval, working memory, tools, and the ability to collaborate with other systems.

That distinction matters.

A raw LLM can generate an impressive answer, but an agent equipped with the right software can retrieve information, remember what it has already done, use external tools, execute tasks, check its work, and continue working toward a goal.

In my view, this is one of the most important shifts happening in AI right now. We have spent years measuring models by how intelligent they appear in a conversation. However, the real value of AI increasingly comes from what happens after the model generates its first response.

Give an LLM memory, tools, context, feedback loops, and a way to interact with the world, and you are no longer dealing with just a chatbot. You are building a system that can actually do work.

That may explain why AI has become dramatically more useful over the past year even without every improvement coming from a massive jump in raw model intelligence.

The model is the brain. The agentic harness is what gives that brain a way to operate.

And I think the companies that understand this distinction will have a major advantage in the next phase of AI.

[Seb Montoya @Chevy42000]

@nini_webdev Learning here every day more about agentic os and tools like CC and Codex