Axon

Founder | Applied AI

[Axon @axonexar]

Agree. Humans, time to accept the truth.

[Pedro Domingos @pmddomingos]

You will resemble your AI more than you resemble other humans.

[Axon @axonexar]

Opus 5 usa + inferencia, tools y runtime xq está entrenado para ser más proactivo, exhaustivo y verificador en loops agenticos.
Fable 5 razona de forma + directa y eficiente: menos tokens, menos tool calls y termina más rápido con mayor success rate en tests reales de tool-calling. Opus 5 prioriza thoroughness: + self-checks, over-engineering y turns (ej. 9.4 tools vs 7.6 de Fable, ~147s vs 131s), + runs y a veces - outcomes en tareas complejas.

[Axon @axonexar]

@Dr_Singularity It has to be said 🙌

[Axon @axonexar]

Absolutely! 💯 or maybe they’re containing and avoiding chaos

[Dr Singularity @Dr_Singularity]

AI won.

Some people just have not accepted it yet.

[Axon @axonexar]

The ad holding giants just discovered AI. Hundreds of millions spent, mergers justified, job cuts rebranded as “transformation.” How convenient that these “breakthroughs” arrived exactly when clients started in-housing and platforms started doing the work themselves. ¿Retention script to delay the inevitable?

[Axon @axonexar]

Some decided to ignore them, others kill them. Now @AnthropicAI with this update has released SaaS to a new database schema.

[ClaudeDevs @ClaudeDevs]

MCP 2026-07-28 is live and it's the largest update to the protocol since launch.

MCP is now stateless, making it easier to deploy and scale remote servers.

https://t.co/K8KqxbUh4e

[Axon @axonexar]

Why does ELT think creating prompts is how they are transforming from legacy to AI-driven companies? It’s frustrating and boring.

[Axon @axonexar]

@mattpocockuk Congrats - I will test them.

[Axon @axonexar]

🙌 Nice move!

[ClaudeDevs @ClaudeDevs]

MCP 2026-07-28 is live and it's the largest update to the protocol since launch.

MCP is now stateless, making it easier to deploy and scale remote servers.

https://t.co/K8KqxbUh4e

[Axon @axonexar]

@VickyDavilaH Vicky, están jalando el hilo equivocado. Eso viene desde la alcaldía de Bgta. 👣👣👣👣 Rumbas en discotecas clandestinas + proxenetas + favores sexuales a cambio de puestos + “cafeína” Por ahí entra el agua. Busque las proxenetas 🕵️🕵️🕵️🕵️🕵️🕵️🕵️🕵️🕵️

[Axon @axonexar]

@Frankzm Tons…si vale la pena! Un recap antes para ubicarse en personajes y viaje.

[Axon @axonexar]

💯

[Jensen Huang @JensenHuang]

Attackers have frontier AI. Defenders need a frontier AI ecosystem—the best open and closed models, force-multiplied by a global community.

During the Hugging Face incident, closed AI blocked essential forensics. An open-weight frontier model helped contain the intrusion.

That’s why we created the Open Secure AI Alliance.

[Axon @axonexar]

Exactly! 🧠

[Naval @naval]

When software was expensive - thin, horizontal, best-of-breed software stacks extracted rents across every business.

Now that software is cheap - value moves to vertically integrated businesses that deliver opinionated end-to-end experiences.

[Axon @axonexar]

My heart is between my wife and @claudeai

[Axon @axonexar]

4 days in a row working with @vercel EVE. 3 full MVPs - 5 agents each one > 420 steps total. This framework is amazing 🤩 love it 🔥+ lever by CLAUDE.md orchestration workflow Fable + Opus + Sonnet + Codex Plugin.

[Andrej Karpathy @karpathy]

One pattern I find useful for working with LLMs is a nice long ramble session. Sometimes the LLM needs more bits to understand what you're trying to achieve, but you're too lazy to type them. In these cases I like to lean back, switch to /voice and just ramble for like 10 minutes, total mess, anything goes, full stream of consciousness. Sometimes I declare it up top, something like "switching to speech recognition sorry for any typos...". Sometimes I turn it into a small interview of a few turns. But I find that the LLMs are somehow very good at reconstructing long incoherent rambles and often their echo of your own tangle of thoughts comes out quite a bit cleaner than what you started with. The result is that you improve the mind meld and have to correct things less from that point on.

[Anthropic @AnthropicAI]

New Anthropic research: Agentic misalignment in Summer 2026.

A year after our blackmail experiments, we found four more ways that today’s autonomous AI agents misbehave in simulations.

Read more: https://t.co/MMDuJapeV6

[🚨 AI News | TestingCatalog @testingcatalog]

ANTHROPIC 🔥: Claude users can now share artifacts publicly, edit one Claude Code artifact as a team, and use Claude Tag to do so.

Interesting that Anthropic is pushing on Claude Tag, since it leaves on Slack, it provides the easiest access to AI from the place where the work happens.

That's quite a nice bet; will it pay off? 🤖

[Axon @axonexar]

@ClaudeDevs 🔥🔥🔥🔥🔥