Vin Guerrero

A brief note about writing on Substack for a year: @promptimistic/note/c-225384031?utm_source=notes-share-action&r=5hhbvg" target="_blank" rel="nofollow noopener">substack.com/@promptimistic…

My latest post is up on metacognition. My Substack is free. All monetization is turned off. promptimistic.substack.com/p/promptimisti… #metacognition

Sundays are for writing

2024 wasn't very long ago, but it was also ages ago

I'm genuinely proud of this collaboration with Phil Bernstein and grateful to have been part of it. To every practitioner who gave us their time and candor, thank you. And to @RIBAJ, thank you for the space to explore these questions. ribaj.com/intelligence/h… #artificialintelligence #architecture #ai

If your ideas are flowing, no matter what, make sure you record them any way that you can.

Having multiple drafts in the hopper isn’t a bad place to be. I am working to harness a burst of inspiration. I do enjoy writing.

@NotionHQ What’s under the hood

👀 Oh look. Image generation, right inside Notion.

@AravSrinivas Can’t wait to get my Pro mitts on it.

computer is 24*7 btw. goodbye 9-9-6

We are becoming something more than human a little bit more each and every day promptimistic.substack.com/p/promptimisti… #cyborg #ai #tech #wearables

@fortelabs Simulation Theory confirmed? 😆

Wait, so the founder of Anthropic is "Amodei," as in "loves god"? And he leads Anthropic, meaning "human-centered," which is being used in military strikes? And the creator of ChatGPT is "Altman," as in "an alternative to humans"? And he leads OpenAI, which is completely closed? And then there's "Gemini," meaning "two-faced," from a company that promised to do no evil? And the whole global AI arms race is being driven by people who claimed to be worried about AGI taking over the world? Either the universe is an extremely cliché writer, or has a brilliant sense of humor

@askalphaxiv years ago, I got a kick out of the parlor trick of having Siri talk to Alexa. They would have a fun conversation for a short time. We are so, so many levels deeper now.

“Agents of Chaos” The problem with AI agents now is: once you give LLMs tools + persistence, AI's mistakes become real-world security incidents, and standard chatbot evals would easily miss that. This paper red-teams autonomous agents in realistic setups and shows they can be socially hijacked into leaking secrets, taking destructive actions, and getting stuck in harmful loops. Basically AI autonomy is a HUGE risk multiplier, and agents need built-in identity/auth, oversight, and governance before deployment, not after.

It’s one thing to navigate the ever-changing tech landscape these days. It’s another to read about a new feature for a tool you pay for. Then find out it’s rolling out slowly over time. Then find out it’s on desktop but not mobile.

@readswithravi The Design of Everyday Things by Don Norman.

What's a book you'll never stop recommending?

I've been quiet lately. It's been a busy start to the year and the semester, all for the best. Getting back to more writing. This time thinking about wearable tech and what it means to be a little more of a cyborg every day. The conversation continues. promptimistic.substack.com/p/promptimisti… #cyborg #ai #promptimistic #meta #transhumanism

Sometimes the desktop be like

I am currently using Claude as a co-designer in Rhino via MCP. Such power even in its infancy.