Jose from Automato

The idea always sounds more useful than the reality. I over-engineered most of my ideas for apps and workflows with AI. Added cool functionality and integrations I actually don't use. Is this really useful, or just exciting to build? https://substack.com/@joseantoniomorales/note/c-276566837

Exactly! Here's the flow for it.

Some people learn AI tools to stay current. Others learn AI to develop a different way of thinking. What they learn might be obsolete in months. They do it anyway — not for the knowledge, but for the instinct it builds. And it compounds. https://substack.com/@joseantoniomorales/note/c-230090026

AI is everywhere new content is created — we can feel the standardization spreading. Same as when filters took over Instagram photos. How are we gonna keep our art when everyone rushes to be standard? https://substack.com/@joseantoniomorales/note/c-265104548

I take Claude similarly to working with a human assistant, imperfect, well-intentioned, and always improving. But now more than ever, Claude helps me learn and build software and automations that I would not be able to do by myself so efficiently.

A new app helps EU funding applicants write better proposals. The same app scores those proposals for the evaluators. When the coach and the judge are the same system — what exactly is being measured? https://substack.com/profile/17146826-jose-antonio-morales/note/c-265851491

Today I saved half the time I usually spend preparing for a client workshop. All my articles, drafts, and presentations in one place, retrieved with Claude's assistance. I just don't have such a memory. https://automato.substack.com/p/ai-assisted-knowledge-management-4u

What should AI do and what should Humans do? There is a line that AI can't cross: Human skill and mastery. It includes more than knowledge and speed; it comes with a deep understanding of Human emotions and behaviors, and a different sense of context.

Software distribution is no longer what most of us assume. We won't buy software much longer. We'll buy instructions — specifications for AI to build exactly what we need, when we need it. https://substack.com/@joseantoniomorales/note/c-240893735

The idea of a second brain sounds cool and a bit science fiction... but I won't let naive enthusiasm make me follow the narrative just like that. My first question is: how can this be useful to me? The discarded question: what can I do with these tools?

For me, context is the real bottleneck — not prompting or tools. AI is a processor more than a thinker, and without clear, structured context, even the best model gives inconsistent results. The judgment still has to come from us.

Can you imagine a day when websites are no longer needed? Today, the visitor brings their own LLM. I’m not talking about GEO, making old websites visible to AI. I’m talking about sites written for the visitor’s LLM to read. https://substack.com/@joseantoniomorales/note/c-357577987

Own your business and your creative process first. Depending on a platform is a high-risk strategy. Build something real — offline, grounded, yours — and let the digital version grow from that foundation. https://substack.com/@joseantoniomorales/note/c-269753391

Why are my team members holding on to information about their jobs? I believe they are confusing information and skill. The information belongs to the company, but their skill is irreplaceable and unique to each one of them. https://substack.com/@joseantoniomorales/note/c-319364598

AI productivity is moving from the IT department to every desk. The companies gaining ground aren’t the ones with the fanciest enterprise AI tools. They’re the ones where individual team members can automate their own workflows - safely, without disrupting existing processes.

Traditional IT is deterministic. Same input, same output. Always. AI is probabilistic. Context shapes the result. That's not a bug. That's the design. Expecting one to behave like the other is the mistake. https://substack.com/profile/17146826-jose-antonio-morales/note/c-236746989

When I subscribe to Claude or ChatGPT, I'm essentially renting processing power that lives somewhere else. My device is just the interface. We keep oscillating between centralized and distributed. Every generation thinks it invented the new model. It never does.

Obsidian and Claude are improving the way I work. Most of my work will be stored as Markdown documents. Claude is already generating the files, adding fields that carry meaning, linking each document to the others that are relevant... automatically. Your computer file system becomes Wikipedia.

How to use AI to improve my work? The low-hanging fruit is to analyze your routines and find manual work that could be automated.

Without process clarity, AI is not a solution. It is an accelerant. If you know your processes, it becomes powerful. If you do not know your processes, it amplifies the confusion. https://automato.substack.com/p/digital-transformation-case-study-for-production-companies

Clients do not pay me for my technical knowledge. They pay for the way I read an organization: where the real friction is. Not the friction people describe, but the friction that actually costs them. https://automato.substack.com/p/your-domain-knowledge-is-the-product

A flow diagram is the one artifact that survives a migration. It doesn't care what platform you're on. It doesn't get deprecated. Anyone on your team — or your future self — can read it and understand what the system does in minutes. automato.substack.com/p/i-built-a-custom-gpt-that-makes-my

If human cognition stops being scarce, what happens to every system we built around it? Gates and Klein circled that question for an hour. A fly in the room had a few thoughts.

I'm working on my relationship with models. I don't want to be dependent on any single one. I'd rather relate to local ones than foreign ones. Obsidian just keeps the whole story: every conversation, every model, always there. No internet required.

We often believe AI thinks like humans because it speaks like one. But AI thinks like AI, and what we're doing now is learning to communicate. https://automato.substack.com/p/ai-doesnt-think-like-you

What AI is exposing is not a lack of talent, but a lack of organizational imagination: rigid processes, slow decision-making, and ineffective structures. https://automato.substack.com/p/ai-2026-the-challenge-is-organizational

Any tool, app, skill, or automation we build with AI should be easy to transfer to other models. We should avoid getting locked into any single LLM. Build your AI tools, for your business and for yourself, with independence in mind. https://automato.substack.com/p/lets-own-our-ai-stack

Give to Claude what is Claude's. I kept asking Claude to fetch the news and drive a browser. Slow, fragile, costly. The fix wasn't a better prompt. It was a better tool next to Claude. Part 1: Apple Shortcuts https://automato.substack.com/p/power-up-claude-ai-1of3-apple-shortcuts

Join us this Wednesday for a special workshop: Esolangs for Artists with @dtemkin.bsky.social . Discover where code meets art and explore the colorful world of esoteric languages. No prior knowledge needed! The event will be in English. Don't miss out! zurl.co/jBGp9

Knowledge belongs to the company, and now it is cheap to get, but personal skills are so much more valuable and belong to the people. https://automato.substack.com/p/ai-enshittify-business

OpenAI builds impressive things, but it's growing very fast. How can we fully trust its competence with so many moving parts? And then there's the need for funding. These companies go beyond ambition, and that creates blind spots.

Open weight isn't a gift. It's leverage, held by someone else, until you understand it well enough to hold some of it yourself. https://automato.substack.com/p/ai-llms-open-source-and-digital-sovereignty

I really like coffee.

Be suspicious of the narratives

What if AI assumes Trump's way of thinking? growingfearless.substack.com/p/if-ai-beco...

You won’t guess it, but I have to tell you: the best place in the world to live is called Slovenia

Be useful and valuable to real people. Digital Audiences Are more like Imaginary ones.

How to grow your confidence level? open.substack.com/pub/growingf...

Apple is suggesting a huge change of direction for AI - LLM models. Very important for AI Strategists and analysts - plus the Claude Fable 5 fiasco… pure and solid validation open.substack.com/pub/automato...

Hoy vota el Perú. Que gane el mejor

That’s my office lately 😍

Can you hear the woodpecker? Peace! ✌️

I can't believe that I found the connection between Matrix, Cambridge Analytica, and the Gnostics! :) open.substack.com/pub/growingf...

You get the options. You decide. Who gives you the options?

This one just got me 60 new subscribers!

A week of errors and Claude limitations... but now you get the prompts and lessons learned. automato.substack.com/p/build-task...