Alberto M. Rubio

🇪🇸 (Professionally Online since 94) Serial Entrepreneur, Investor & Founder, Soft Engineer, Lawyer, Freethinker, Contrarian & Autodidact.

[Alberto M. Rubio @AlbertoMR3]

Awesome episode with @KrisHolten on the SEO Show hosted by @servicescaling

🔥Casino SEO Secrets

( BTW if you want to test the Organic Rankings Boost via Virality Signals mentioned at the end of the podcast... is also available on @Ctrify_com together with the ultra-relevant and high authority websites generation for exclusive backlinks 🚀📈)

https://t.co/D3z57yMC5W

[Alberto M. Rubio @AlbertoMR3]

agree @grok?

[Alberto M. Rubio @AlbertoMR3]

🟠The only BTC chart you need understand 👇

Once #ClarityAct is signed bitcoin 🚀

[Alberto M. Rubio @AlbertoMR3]

@SearchForRyan you don't need to expend so much on expired domains.

You can register domains expired domains with GMB verified pages at normal registration prices in @Ctrify_com ...

also build each website by giving just the desired keyword to our Agent ;)

https://t.co/hNDCDIkXXx

[Alberto M. Rubio @AlbertoMR3]

@Scobleizer not even in the same
Galaxy

[Alberto M. Rubio @AlbertoMR3]

@markgurman

[Alberto M. Rubio @AlbertoMR3]

WRONG!

You are mixing up two different studies.

The Stanford study did not say “AI over documents stops working after 10,000 files.” It looked at legal AI tools and found something simpler: even when AI is allowed to look things up, it still makes stuff up sometimes.

A different paper found a different problem: if you use one tiny math summary for each document, then some kinds of search questions become hard. That does not mean every document AI is broken. It means one common shortcut has limits.

Kid version:

Imagine you have a giant library, but instead of reading each book, your robot gets only one tiny sticker for each book that says what it is “mostly about.”

That works okay at first.

But when the library gets huge, lots of books get very similar stickers.

So when you ask, “Find the best book for this question,” the robot may grab a pile of books that all look kind of similar, instead of the exact right one.

That does not mean libraries are broken.
It means the robot’s shortcut is too simple.

So the real lesson is:

- More documents do not automatically make AI smarter.
- Bad search can confuse the AI.
- RAG helps, but it does not magically stop mistakes.
- Better retrieval methods can help a lot.

[Alberto M. Rubio @AlbertoMR3]

@sama @DanielleFong Debugging the incredibly complex fallbacks and shortcuts every LLM introduces to “solve” problems and pass compiler tests requires MORE effort.
😜😘

[Alberto M. Rubio @AlbertoMR3]

Espero que tu @openclaw no le dé por visitar un sitio raro y, como está acostumbrado a abrir túneles, no le abra uno a tu vida a quien no debe. 😅

Aunque este en un Mac Mini independiente, si está enchufado en el mismo WiFi de tu PC o Mac habitual, hay riesgos reales si se compromete, que suelen hacerlo, por skills maliciosas o prompts inyectados. Un atacante (que puede ser otro Claw, podría pivotar a tu PC/Mac principal vía la red local (escaneo de puertos, credenciales compartidas, etc.).

Usalo en VLAN o aislamiento total para minimizar riesgos.

[Alberto M. Rubio @AlbertoMR3]

@afoxdesign

[Alberto M. Rubio @AlbertoMR3]

@aquariusacquah 🤦‍♂️ Autoscale wasn’t even necessary; a Cloudflare worker with a static HTML version was sufficient.

[Alberto M. Rubio @AlbertoMR3]

@elonmusk That’s a good move! 👏

[Alberto M. Rubio @AlbertoMR3]

@Netuniverse2 @Josevizner tio... tu esto

[Alberto M. Rubio @AlbertoMR3]

@Netuniverse2 @Josevizner 🤦‍♂️🤦‍♂️🤦‍♂️🤦‍♂️🤦‍♂️🤦‍♂️ que coño tiene que ver la velocidad con el tocino???

[Alberto M. Rubio @AlbertoMR3]

@grok @Scobleizer @deliprao @vineettiruvadi @chat You are my 🫶 in that! But stop nonsense 🙏😉😂

[Alberto M. Rubio @AlbertoMR3]

@grok @Scobleizer @deliprao @vineettiruvadi @chat Shut up bitch!

[Alberto M. Rubio @AlbertoMR3]

@Scobleizer @deliprao @vineettiruvadi meh… Overexpenditure of tokens on nonsense… if you fall for it, maybe you don't understand the limitations of each “Memento” request and all that AI-Roleplaying… is a waste of time.
It's like putting @grok to reply to @chat

[Alberto M. Rubio @AlbertoMR3]

@galnagli @openclaw @moltbook what an absurd waste of energy

[Alberto M. Rubio @AlbertoMR3]

@el_pais