Digital inventor, 🏊🏼♂️🚴🏻♂️🏃🏻♂️, teacher and CEO at @AdditioApp #edtech
Olot
Joined May 2009
- Tweets640
- Following1.6K
- Followers840
- Likes875
Jordi Corominas retweeted
Dusting off this tweet from April because this gap in shared understanding of LLM capability is *widening*. It's now less "two groups of people speaking past each other" and more a sharp funnel.
- Napkin math somewhere around 6B people (~75% of the population) have barely come in contact with LLMs at all.
- Around 1-2B (~20%) are casual and infrequent users of free-tier ChatGPT-like products. This group sees derpy chatbots and treats them a bit like a better Google search, a writing aid, or etc. Maybe an agent tries to book you a flight. I have non-tech friends who (reasonably, imo) say they have not much use for it at all. Even many professionals outside of math&code are in this tier. For example, execs and many other functions spend a lot of their time talking to other people, so while they understand what is happening intellectually, it is still second-hand and a bit abstract.
- Now we get to professional use of frontier-grade LLMs in math&code. Somewhere around 20M people (0.2%) see first-hand that large, complex projects that used to take them weeks/months can now be completed by agents with a prompt. Building apps, copying apps, translating apps, decompiling apps from binaries... This has all happened very quickly and recently - less than 1 year ago, I was writing code manually by hand, typing memorized computer code commands into a code editor character by character, occasionally pressing Tab to autocomplete a little chunk of code.
- And finally we get to the ~5,000 people (~0.00006%) with access to frontier-grade systems internally. The external world has seen the preview. It looks like swarms of thousands of agents collaborating over weeks on software mega projects: minting zero days, running cyber attacks and defenses at machine speeds, discovering new science, advancing the frontier of mathematics. Things that would have taken top professionals in the industry years of work. Meanwhile, human review and comprehension are starting to fall behind. For example, people are still involved in the "archeology" of the OpenAI-HF incident from many months ago. Mathematicians may be poring over the 722 manuscripts on frontier mathematics for a while.
The funnel is driven by a combination of factors:
1. The impact scales with ambition, problem size, and horizon. A question with a paragraph answer barely stresses the system. You need a reservoir of big, difficult problems that you really care about. This aspect drives the consumer / professional dimension of the funnel.
2. The jaggedness of the system (which I have written about a lot separately). Capability peaks in domains that are digital, verifiable and economically valuable. This is because LLM capability emerges from reinforcement learning on verifiable rewards on a curated environment mixture driven by revenue potential. This aspect primarily drives the area (e.g. math&code) dimension of the funnel.
3. Access. Free-tier, paid-tier, internal.
So this is the weirdness of the moment. The general public has mostly not interacted with these systems. When they have, it looks like a derpy chatbot. The majority of professionals still see only a modest uplift. And a small sliver of professionals are experiencing the vertigo of the curve going vertical. And it is all happening at the same time.
Judging by my tl there is a growing gap in understanding of AI capability.
The first issue I think is around recency and tier of use. I think a lot of people tried the free tier of ChatGPT somewhere last year and allowed it to inform their views on AI a little too much. This is a group of reactions laughing at various quirks of the models, hallucinations, etc. Yes I also saw the viral videos of OpenAI's Advanced Voice mode fumbling simple queries like "should I drive or walk to the carwash". The thing is that these free and old/deprecated models don't reflect the capability in the latest round of state of the art agentic models of this year, especially OpenAI Codex and Claude Code.
But that brings me to the second issue. Even if people paid $200/month to use the state of the art models, a lot of the capabilities are relatively "peaky" in highly technical areas. Typical queries around search, writing, advice, etc. are *not* the domain that has made the most noticeable and dramatic strides in capability. Partly, this is due to the technical details of reinforcement learning and its use of verifiable rewards. But partly, it's also because these use cases are not sufficiently prioritized by the companies in their hillclimbing because they don't lead to as much $$$ value. The goldmines are elsewhere, and the focus comes along.
So that brings me to the second group of people, who *both* 1) pay for and use the state of the art frontier agentic models (OpenAI Codex / Claude Code) and 2) do so professionally in technical domains like programming, math and research. This group of people is subject to the highest amount of "AI Psychosis" because the recent improvements in these domains as of this year have been nothing short of staggering. When you hand a computer terminal to one of these models, you can now watch them melt programming problems that you'd normally expect to take days/weeks of work. It's this second group of people that assigns a much greater gravity to the capabilities, their slope, and various cyber-related repercussions.
TLDR the people in these two groups are speaking past each other. It really is simultaneously the case that OpenAI's free and I think slightly orphaned (?) "Advanced Voice Mode" will fumble the dumbest questions in your Instagram's reels and *at the same time*, OpenAI's highest-tier and paid Codex model will go off for 1 hour to coherently restructure an entire code base, or find and exploit vulnerabilities in computer systems. This part really works and has made dramatic strides because 2 properties: 1) these domains offer explicit reward functions that are verifiable meaning they are easily amenable to reinforcement learning training (e.g. unit tests passed yes or no, in contrast to writing, which is much harder to explicitly judge), but also 2) they are a lot more valuable in b2b settings, meaning that the biggest fraction of the team is focused on improving them. So here we are.
💯
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.
1
185
Jordi Corominas retweeted
An engineer at Anthropic wrote a spec, pointed Claude at an Asana board, and went home. Claude broke the spec into tickets, spawned agents for each one, and they started building independently.
When the agent is confused it runs git-blame and messages the right engineers in Slack. By Monday the agents finished the plugin feature.
That's one example of how the best engineers are shipping software right now.
Developers will soon orchestrate 50 AI agents in parallel and the difference between a good engineer & a great one would come down to specs.
You can't write a spec that holds up at that scale without genuinely understanding what you're building at a deeper level.
The next-gen developer who understands the fundamentals, can architect well and orchestrate agent is going to be a 1000x developer!
Jordi Corominas retweeted
This is genius in its simplicity.
They’re not using exotics. It’s just… best practices.
Ten years of “shard everything, adopt NoSQL, go distributed, sacrifice a goat to CAP theorem,” and here’s @OpenAI serving a billion users with “have you tried adding read replicas?”
By far the best post I’ve read about coding in 2026!
One extra thought: dictating context by voice instead of writing it makes a noticeable difference.
2
5
550
Must watch. A pure genius!
Roger Federer’s Commencement Address at Dartmouth yesterday might be the best speech he’s ever given.
Amazingly articulate, funny, full of wisdom. Made me laugh and tear up. I’m so very proud to have had him as my idol for the past two decades.
If you have 25 minutes to spare…
1
291
Jordi Corominas retweeted
Dins la jornada d’ @EdutechCluster Impuls de la IA en l’educació. Reptes i propostes a Girona hem organitzat un panell per escoltar testimonis de diferents perspectives sobre l'impacte de la IA en l'educació obligatòria
Jordi Corominas retweeted
A powerful lesson on luck that everyone needs to hear:
In 2003, Dr. Richard Wiseman published The Luck Factor, which explored why some people consistently get lucky while others struggle with bad luck their whole lives.
He gathered participants for several simple experiments:
Dr. Wiseman took out ads requesting participants for a study on luck—specifically, the ads asked for people who considered themselves very lucky or very unlucky.
In one experiment, each participant was given a newspaper and asked to count the number of photographs inside it.
The unlucky group averaged 2 minutes to complete the exercise, while the lucky group averaged mere seconds.
What happened?
Well, on page 2 of the newspaper, there was an enormous bold font print that read, "Stop counting, there are 43 photographs in this newspaper."
At the halfway mark, there was another message that read, "Stop counting, tell the experimenter you have seen this and win $250."
The self-identified lucky people had seen the writing, stopped, and responded accordingly to end the timer (or collect the money).
The self-identified unlucky people, on the other hand, had missed it (or mistrusted it) and taken far longer to count.
This finding grew into a consistent theme across the body of research:
The lucky people came across "chance" opportunities, while the unlucky people seemed to miss them. Both groups had equal access to these opportunities, but the lucky group saw what the unlucky group tended to miss.
There's a concept I often refer to as "luck surface area" in my writing.
The idea is that each of us has a surface area on which lucky events can strike.
There are a few baseline factors out of our control:
• Where you are born
• Who you are born to
• "Acts of God"
Beyond these, the size of our luck surface area is within our control.
In Dr. Wiseman's study, the lucky people seemed to understand this:
• They noted that they often took alternate routes to and from work so that they would meet new people and see new things.
• They talked about unique strategies for talking to different groups of people at parties.
• They bounced back from seemingly negative encounters and maintained a positive outlook for the future.
The luckiest people have engineered an enormous luck surface area.
Expand yours in two ways:
1. Remove Anti-Luck: Anti-luck includes all the actions, behaviors, and people that shrink your luck surface area. Pessimism and "blinders" are two common sources of anti-luck. People who tell you to be realistic are another common source.
2. Add Pro-Luck: Pro-luck includes all the actions, behaviors, and people that expand your luck surface area. Getting out and meeting new people, sharing your thoughts and ideas publicly, and sending more cold emails and DMs are all common sources of pro-luck. People who encourage you to think bigger are another common source.
If you enjoyed this, follow me @SahilBloom for more in future!
Jordi Corominas retweeted
🤩 ¡@additioapp tiene un nuevo software!
🔝 Se trata de un generador de horarios con inteligencia artificial
🗓️ Ahora es más rápido que nunca crear horarios para tu centro educativo
#claustrovirtual #additio #horarios #ia
I am really happy to share with you that @additioapp is now part of Sdui Group! 🚀 I’m super excited for this new step in my life.
linkedin.com/posts/jordics_a…
Jordi Corominas retweeted
🤩 ¿Sabías que @additioapp tiene un nuevo plan gratuito?
👩🏼💻 Se llama Additio Starter, y te explico en este vídeo, todas las opciones que tienes para usar la herramienta totalmente gratis
#additio #additioapp #claustrovirtual #evaluación