{"slug": "human-bottlenecks", "title": "Human Bottlenecks", "summary": "AI models are increasingly capable, but their potential remains underutilized due to two main bottlenecks: lack of a serious context of use and internal human factors. Most people who desire AI tools like flashcard generators or tutors do not have a genuine need for them, and even advanced AI cannot overcome the absence of motivation or a concrete problem to solve.", "body_md": "AI models are very capable and get more capable each year. So naturally people feel they’re underusing them. There’s a tweet that goes like: your laptop has a 100M USD startup in it, you just have to figure the right sequence of words to get it out. And beyond money, people imagine AI could boost them in every area of life. Thus all these perennial ideas: of an AI executive assistant, an AI tutor, an AI that curates your “digital garden”, an AI that (sigh) writes flashcards for you.\n\nThe general template is: if only I could wire up the right prompts and the right\ntools in the right harness, I could have an agent that would boost my\nproductivity 10x, or fix my problems with therapy, or make me more social, or\nmore knowledgeable. This was, curiously, the ambition of a lot of early\ncomputing pioneers: [ Augmenting Human Intellect](https://www.dougengelbart.org/pubs/augment-3906.html),\n\n[.](https://en.wikipedia.org/wiki/Man%E2%80%93Computer_Symbiosis)\n\n*Man-Computer Symbiosis*[Engelbart](https://en.wikipedia.org/wiki/Douglas_Engelbart)’s lab was called the\n\n[Augmentation Research Center](https://en.wikipedia.org/wiki/Augmentation_Research_Center)! And more recently, people used to complain about how everyone has the Library of Alexandria in their pocket, and yet, we are not all genius polymaths.\n\nAnd these ideas are perennial because they never seem to happen. It’s like the\n[Solow paradox](https://en.wikipedia.org/wiki/Productivity_paradox) on an individual level. Why? I think there are two\nreasons: first, most people lack what [Andy Matuschak](https://andymatuschak.org/) calls a “[serious\ncontext of use](https://notes.andymatuschak.org/People_who_write_extensively_about_note-writing_rarely_have_a_serious_context_of_use)” (AI doesn’t move the needle because there’s no needle to\nmove); second, most people are bottlenecked by internal factors where AI (or\nanything external, for that matter) can’t move the needle.\n\n# The Serious Context of Use\n\nI have heard so many people, online and in real life, tell me some variation of:\n“I want to [use|build] an app that uses AI to write flashcards”. How many of\nthose people do you think have ever written a flashcard? How many of them use\n[Anki](https://apps.ankiweb.net/) every day? Have ever used Anki? The people who want an AI that\nwrites flashcards for them *don’t use flashcards*. They have no reason\nto. Dually, the people who use flashcards [would benefit little from AI](/article/the-applicability-of-spaced-repetition)\nwriting their flashcards.\n\nAnalogously with “AI tutors”. If you had the ghost of John von Neumann in your\nlaptop, what would you have him teach you? Let’s be honest. You’d go through\nchapter one of some math topic you’re vaguely curious about and then forget\nabout it. And that would probably be the rational move! Most people are not\nautodidacts because most people have no material reason to learn a specific\ntopic (i.e. their job does not require it) and the problem with learning for the\nsake of learning is opportunity cost: there is no *a priori* reason to learn one\nthing over another, so better to do nothing and wait for something to appear\nwhich actually grabs your interest. Again, this is likely rational! Could you\nimagine if you found everything interesting? You’d spend years living in a\nbasement curating a wiki of late Soviet military hardware or something. So, even\nif you had John von Neumann in your pocket, it probably wouldn’t move the\nneedle.\n\nWould an “AI executive assistant” *actually* boost your productivity? What would\nit do, other than tell you to do the things you already know you have to do?\nWith these ideas that are so attractive in the abstract, the way you deflate\nthem is you interrogate the concrete, fine-grained details. Take a day at work,\nand ask: what *exact* actions could an AI looking over my shoulder have taken,\nthat would have made a difference?\n\nFinally there’s the tools-for-thought/notetaking people. God save us. It’s\nalways the same thing. Your folder with notes—pardon me, your “[second\nbrain](https://fortelabs.com/blog/basboverview/)”—plus an AI agent that writes, edits, synthesizes information,\nanswers queries. You could build this in an afternoon, and it won’t move the\nneedle in your life, for the same reason that building the second brain in the\nfirst place didn’t make a difference.\n\nSee, most of us, unless we are students, we really don’t have cause to take notes on anything. If you’re a student, you take notes from the textbook. I keep a journal, which is occasionally useful. At some jobs I’ve kept a work journal, this has also been useful. If I stopped, probably, not much would change.\n\nThe notetaking people—and I say this with all the love in the world—are\nnever, like, a researcher at the cutting edge of their field, building this vast\ncathedral of knowledge, note-by-note, so they can derive new insights. Never a\nhistorian who has to read tens of millions of words across thousands of sources\nto synthesize the life of some historical person. It’s never someone doing\nsomething hard. It’s always some blogger. Their “digital garden” is about how to\nkeep a digital garden. It’s very solipsistic: there’s no output, no\ndeliverables. The deliverable is you take a screenshot of your [Obsidian](https://obsidian.md/)\ngraph and tweet about it to show off how much it looks like an incomprehensible\nball of twine.\n\nSo, what difference is the AI going to make? “It’s going to write my notes”. About what? “It’s going to read articles for me and summarize them and add them to the digital garden”. For what purpose? “It’s going to find connections between my ideas!” What ideas? It’s going to pull an unfinished list of bulletpoints for an eventual draft of an essay on some inane thing, plus a bunch of PDFs you haven’t read, and combine them together and make, what? Another project you’re not going to do? The AI is going to do that?\n\nAgain, I say this with all the love in the world. I used to be a tools-for-thought guy. I hoard PDFs. But we have to be honest with ourselves. Sometimes, tools don’t move the needle because there’s no needle to move. Because the “needle” is not a concrete, realizable material need but a vague, aspirational idea about who we are as people.\n\n# Internal Limiting Factors\n\nSo, the idea of using computers to augment human capabilities is basically: you\ntake the human, and you build a scaffold around them, but the human stays the\nsame. The scaffold can be classical software, or AI, but the human remains a\nblack box. And the hope is: I just prompt my swarm of AI agents and become 100x\nmore effective, like Manfred in [ Accelerando](https://www.antipope.org/charlie/blog-static/fiction/accelerando/accelerando.html).\n\nWhy wouldn’t this work? I think most people are bottlenecked by internal factors that are difficult to change. Mental energy, motivation, executive function, not to mention more fundamental traits like intelligence and conscientiousness. So, external scaffolding, either with classical software or AI, might help somewhat, but it won’t be transformative.\n\nConsider executive function. [My own experience of managing ADHD](/article/notes-on-managing-adhd) is the\nexternal scaffolding helps (todo lists, calendars, timers, a million little ways\nto trick myself into working) gets me from zero to “kind of functional”. But it\nsaturates there. Stimulants fix the original, internal bottleneck, which is my\nneurochemistry. And then I can accomplish my goals (c.f. [Liebig’s law of the\nminimum](https://en.wikipedia.org/wiki/Liebig%27s_law_of_the_minimum)). All the pomodoros in the world are as nothing to a little\nmolecule diffusing through my brain tissue, binding to [NET](https://en.wikipedia.org/wiki/Norepinephrine_transporter) and\n[DAT](https://en.wikipedia.org/wiki/Dopamine_transporter). And what scaffolding is useful is just classical software:\n[Todoist](https://www.todoist.com/) and calendars. Is an agent going to match the effectiveness\nof methylphenidate in ADHD? I doubt it.\n\nConsider intelligence. Can AI augment human intelligence? What does that look\nlike? Consider how AI agents only became useful when the models crossed a\nparticular capability threshold, i.e., you can’t put GPT-2 into a harness and\nget GPT-5 outcomes out of it. Can you put a human in an AI scaffold and give\nthem +30 effective IQ? I doubt it, unless the AI is doing all the thinking, at\nwhich point, what use is the human? The limiting factor in the human-AI\n[centaur](https://en.wikipedia.org/wiki/Advanced_chess) is the human! So intelligence is fixed until we get very\nadvanced biotechnology.\n\nKnowledge is another limiting factor. I find that even very educated people tend\nto underrate the importance of knowledge. A lot of people have this attitude\nthat you can just Google everything just-in-time as it comes up. Like Babbage, I\ncan’t [rightly apprehend the confusion of ideas](https://en.wikiquote.org/wiki/Charles_Babbage#Passages_from_the_Life_of_a_Philosopher_(1864)) that would lead someone\nto think this. Maybe it’s downstream of the lack of a serious context of\nuse. Everything you do, every action and idle thought, draws on this vast\n(implicit, unseen) trove of knowledge. Claude Shannon invented digital computing\nbecause he remembered this then-obscure branch of mathematical logic called\nBoolean algebra and saw that it could be realized in hardware. A trillion-dollar\nindustry, conjured out of some old tomes.\n\nThe reason knowledge is still a bottleneck, in the AI era, is not: “if you don’t\nhave the knowledge, you can’t write the prompt”. Rather: if you don’t have the\nknowledge, you don’t understand the question, or why it matters, or how to judge\nthe answers, and you won’t ever think to ask. You’re in a *completely* different\ncontinent from “writing the prompt”. And because long-term memory is private and\ninternal, AI can’t boost it. It can, maybe, with judicious use, help in the\nacquisition of new knowledge.\n\nSo: executive function, intelligence, and knowledge are huge bottlenecks to what\nyou can do, and because they are internal to the brain, AI can’t touch them\nuntil we have far more advanced biotechnology. Corollary: contrary to the\npopular view of [human capital is becoming worthless](https://www.youtube.com/watch?v=2HMPRXstSvQ), the returns to\neducation are now *higher*, because intelligent, educated people with working\nreward circuitry stand to gain more from AI.", "url": "https://wpnews.pro/news/human-bottlenecks", "canonical_source": "https://borretti.me/article/human-bottlenecks", "published_at": "2026-05-18 00:00:00+00:00", "updated_at": "2026-06-13 22:31:46.372576+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-agents", "ai-products", "ai-tools", "ai-research"], "entities": ["Andy Matuschak", "Anki", "John von Neumann", "Doug Engelbart", "Augmentation Research Center", "Solow paradox", "Library of Alexandria"], "alternates": {"html": "https://wpnews.pro/news/human-bottlenecks", "markdown": "https://wpnews.pro/news/human-bottlenecks.md", "text": "https://wpnews.pro/news/human-bottlenecks.txt", "jsonld": "https://wpnews.pro/news/human-bottlenecks.jsonld"}}