Alvaro Rojas

Fan of kind, authentic people.
Founder @hyriseacademy
📧 Newsletter:
Stay curious! 🖖

[Mihura @XMihura]

este domingo no se juega solo una final del mundial: se enfrentan dos formas opuestas de entender la vida

por un lado, Argentina representa la materialización una low trust society. la exaltación de la trampa, el marrullerismo con la excusa de la pasión

un equipo que ha llegado a la final arrastrándose entre ayudas arbitrales y juego sucio, porque en su modelo del mundo el que respeta las reglas es un pardillo

se aprecia en la estética: tíos hiper-tatuados que parecen sacados del patio de una carcel panameña, sin clase, sin saber estar y con un comportamiento visceral sin respeto por el rival

por el otro lado, España representa el orden, la estructura, la táctica. el respeto por el rival y por las normas del juego. La representación de una high trust society y de los valores europeos llevados al césped: un sistema donde cada individuo cumple su función sin hacer trampas, porque confía en que el sistema y el esfuerzo colectivo le van a recompensar

no es solo fútbol, este domingo se enfrenta la civilización contra la barbarie

[Alex Veremeyenko @alex_verem]

The biggest scam humanity has collectively accepted as normal, according to Claude

[Abdulkadir | Cybersecurity @cyber_razz]

“A man who has not hit his Claude limit by noon has wasted his morning.”

-Socrates

[Alvaro Rojas @ajoered]

Three words: Vibe space travel

[SpaceX @SpaceX]

SpaceX has exercised the option to acquire @cursor_ai in an all-stock transaction with the goal of building the world’s most useful AI models.

For the past few months, SpaceXAI has been jointly training a model with Cursor, which will be released in Cursor and Grok Build soon.

We look forward to working closely with the Cursor team to advance our frontier AI capabilities

[Alvaro Rojas @ajoered]

@GuliMoreno @jaimenovoa *Alemanes, eso sí 🇩🇪

[Alvaro Rojas @ajoered]

@GuliMoreno @jaimenovoa Nosotros lo hicimos así. Dos rondas con los mismos BAs los primeros dos años, self-funded growth desde entonces (5 años).

[EL MUNDO @elmundoes]

Aparece un tiburón blanco por sorpresa mientras limpiaban de redes fantasma el mar

🎥 Fundación Healthy Seas

[Alvaro Rojas @ajoered]

This.

[Aaron Levie @levie]

There’s no amount of intelligence that can get packed into AI models that replaces the need for context. For any sufficiently general purpose AI, you will always have to guide it in the direction you want as it has an infinite range of directions it can go in.

As long as the same model is used by a lawyer, an engineer, a financial analyst, or a healthcare professional, and as long as you’re trying to do anything uniquely differentiated or specific, then instructions, domain context, and proprietary data will always need to get into the context window for the model to be useful.

This is partly why AI automation doesn’t come for free, and why there’s still a wide spectrum of who’s getting the largest gains from AI and who’s not. You have to put in real work, and you get real value on the other end.

This is one of the advantages that applied AI will also have in the market. Any layer of abstraction above just the raw intelligence that can meaningfully get you off to the races faster will likely continue to be valuable.

[juanmacias 🏳️‍🌈 @juanmacias]

Consejos para definir procesos en una empresa no nueva (Vamos, poner un poco de orden).

Llevo definiendo procesos más de 20 años, y aún me considero un novato, pero comparto algunas cosas que he aprendido:

[monos estocásticos @monospodcast]

La lectura del día, del mes y ya veremos si del año:

https://t.co/2uz4uWI02D

[Alvaro Rojas @ajoered]

I've been building with Claude Code almost every day for months, and one of my biggest learnings is how crucial context/memory management is.

If you are building some small quick app or use case, you can usually get it to work within one single session. When the build gets complex, the LLM output just gets worse and worse as time and complexity grow.

Seems there is even a name for it now: context rot!

The best solution I've found is having different "memory" files for specific contexts, and one "router" file that points the LLM to each of them. In Claude, the router file is CLAUDE.md and sub files live in .claude/rules/ (one file per domain: architecture, testing, security, etc.)

Every new session, all the important context is already there.

I know you can get way more sophisticated here with compression techniques, sub-agents etc. so interested to see how people are actually managing this.

One of the best resources I've found: https://t.co/mMiH8htT6Y

[Mihura @XMihura]

https://t.co/C8VgHpKVbp

[David Senra @davidsenra]

Dana White on why he doesn’t believe in introspection:

“If you just sit around and talk about your fucking problems all the time it actually makes it worse.

I never take in any negativity.

I literally block it out.

I block all the noise out.

Like these guys who report on what we're doing that have no clue on what we're doing?

Why would I want to hear anything they have to say?

They're zeroes.

They've literally never done anything in their life, especially in this business.

Why would I listen to anything that they have to say?”

CC @pmarca

[Miguel Carranza @elwatto]

insane gains. save a rainforest and your budget https://t.co/YPhVE3ACAr

[David Álvarez de la Torre @davidalvarezdlt]

Las startups del futuro:

- 1 backend
- 1 front
- 1 infra (opcional)
- 1 marketer
- 50 customer sucesss
- 125 vendedores

[Aaron Levie @levie]

Another week on the road meeting with a couple dozen IT and AI leaders from large enterprises across banking, media, retail, healthcare, consulting, tech, and sports, to discuss agents in the enterprise.

Some quick takeaways:

* Clear that we’re moving from chat era of AI to agents that use tools, process data, and start to execute real work in the enterprise. Complementing this, enterprises are often evolving from “let a thousand flowers bloom” approach to adoption to targeted automation efforts applied to specific areas of work and workflow.

* Change management still will remain one of the biggest topics for enterprises. Most workflows aren’t setup to just drop agents directly in, and enterprises will need a ton of help to drive these efforts (both internally and from partners). One company has a head of AI in every business unit that roles up to a central team, just to keep all the functions coordinated.

* Tokenmaxxing! Most companies operate with very strict OpEx budgets get locked in for the year ahead, so they’re going through very real trade-off discussions right now on how to budget for tokens. One company recently had an idea for a “shark tank” style way of pitching for compute budget. Others are trying to figure out how to ration compute to the best use-cases internally through some hierarchy of needs (my words not theirs).

* Fixing fragmented and legacy systems remain a huge priority right now. Most enterprises are dealing with decades of either on-prem systems or systems they moved to the cloud but that still haven’t been modernized in any meaningful way. This means agents can’t easily tap into these data sources in a unified way yet, so companies are focused on how they modernize these.

* Most companies are *not* talking about replacing jobs due to agents. The major use-cases for agents are things that the company wasn’t able to do before or couldn’t prioritize. Software upgrades, automating back office processes that were constraining other workflows, processing large amounts of documents to get new business or client insights, and so on. More emphasis on ways to make money vs. cut costs.

* Headless software dominated my conversations. Enterprises need to be able to ensure all of their software works across any set of agents they choose. They will kick out vendors that don’t make this technically or economically easy.

* Clear sense that it can be hard to standardize on anything right now given how fast things are moving. Blessing and a curse of the innovation curve right now - no one wants to get stuck in a paradigm that locks them into the wrong architecture. One other result of this is that companies realize they’re in a multi-agent world, which means that interoperability becomes paramount across systems.

* Unanimous sense that everyone is working more than ever before. AI is not causing anyone to do less work right now, and similar to Silicon Valley people feel their teams are the busiest they’ve ever been.

One final meta observation not called out explicitly. It seems that despite Silicon Valley’s sense that AI has made hard things easy, the most powerful ways to use agents is more “technical” than prior eras of software. Skills, MCP, CLIs, etc. may be simple concepts for tech, but in the real world these are all esoteric concepts that will require technical people to help bring to life in the enterprise.

This both means diffusion will take real work and time, but also everyone’s estimation of engineering jobs is totally off. Engineers may not be “writing” software, but they will certainly be the ones to setup and operate the systems that actually automate most work in the enterprise.

[Alvaro Rojas @ajoered]

El systems thinking sigue siendo la clave, pero parece que el bueno de Mark quiere cambiar el modelo.

La especialización (SDR → AE → CSM) fue una respuesta a las limitaciones humanas, y si la IA elimina esa limitación, el modelo cambia. Lleva tiempo diciendo que la IA reduce la ventaja de especializar roles y que volvemos al full-cycle rep.

O igual solo nos quiere vender el nuevo libro 😏

https://t.co/n06JRawqC9

[Alvaro Rojas @ajoered]

@DiegoARRG DM

[Rory O'Driscoll @rodriscoll]

Intelligence, generated by foundation models like Claude, will infuse all software over the next decade. That’s a given. The question is how it gets to the enterprise. There are five (non-exclusive) paths for this to happen. Enterprises can:

1. Buy directly from the foundation models (the models “just work”)
2. Build it for themselves on top of foundation models (AI for proprietary advantage)
3. Buy from new companies, founded post-2022, built on top of foundation models
4. Buy from existing software vendors who integrate AI into existing apps
5. Buy the outcome from a full-stack, AI-enabled provider (AI lawyer, etc)

Options one or two imply the death of all the non-foundation model application software business (“aka software is dead”). Option three implies all pre-2022 application software is dead or at best legacy (“SaaSpocalypse”). Option four says you buy the WCLD index at 8x EBITDA, and option five has you planning AI-enabled roll-ups.

Figuring out, market by market, how this plays out and why is where the venture alpha is for the next five years.

Hint: answers will vary for reasons that in retrospect will be obvious, and the pace of adoption will differ significantly by market.