Mike Masnick 's Techdirt Comments

Latest Comments (23254) comment rss

  • On the post: My Kid Vibe Coded Their Way To Actually Learning Math

    Mike Masnick ( profile ), 28 May, 2026 @ 08:09am

    Yes. It is a lot of work. And that's kind of my point. Those selling AI as some sort of magic potion that doesn't involve actual work are lying to you. But for those willing to put in some initial work, the benefits are there. Yes, it was a lot of work to initially setup, but once it was working, it continues working. Now it's just pushing a button and getting back very useful results every time. This is the point I keep trying to make. It can be helpful, but not in the way that it's being marketed for the most part.

  • On the post: My Kid Vibe Coded Their Way To Actually Learning Math

    Mike Masnick ( profile ), 27 May, 2026 @ 03:15pm

    Yes, GenAI has its uses and when applied in those limited situations it can be a force multiplier. Just as Mike found it helpful for reviewing his work and others have found similar examples, I too use AI to rapidly review documents and organize data. Despite these limited successes, AI is overhyped, oversold, overpromised, and incapable of fully replacing people.
    And I have consistently said that it's overhyped and oversold by the big AI companies. Not once have I suggested otherwise. I also think that they're all full of shit in claiming that it will replace jobs. I wrote a whole thing on how that was bullshit. If anything it just helps people at the edges of their jobs, and isn't good at all for replacing people. So I agree with you and have written as much. So I honestly don't get this line of complaints. It seems that every single time I say something, where I put directly into context some way in which the tech is helpful a bunch of people start screaming about me about what an asshole I am because I haven't 100% condemned the tech. But I have called out the overhype and bullshit marketing projections. And that's just not good enough for some people? Honestly, the annoying thing here is it makes me want to write about the downsides of AI much, much less because that's so tired. I tried to find a narrow area that I could write about that was neither unbounded hype, nor overhyped doom and gloom. I found a realistic line, and I get yelled at for it as well. So, honestly, that makes me feel like I should fuck off even trying to talk about the negatives since people are going to ignore that anyway. I have to be honest, it really feels like a bunch of people here are desperate to have me lie to feel soothed. It's fucking weird.

  • On the post: My Kid Vibe Coded Their Way To Actually Learning Math

    Mike Masnick ( profile ), 27 May, 2026 @ 01:20pm

    I mean, this is silly for multiple very stupid reasons.

    1. I have put the Pope's thing on my (very long list, over 40 potential stories this week, of which I'll be lucky to cover 5 or 6). I don't know if I'll get to it, but we always tend to cover things a few days later anyway, so I might still get to it. We'll see.

    2. Honestly, I'm not sure what the freaking Pope has to say on tech issues is particularly interesting or relevant to me. He's not exactly known as being relevant on tech policy.

    3. What my kid did was not a powerpoint presentation and I seriously question your comprehension ability if you think that's what happened here.

    4. There are approximately 50k articles about what the Pope said if you want to find one. There is (and can be) only one about what my kid did here. So, yeah, I actually do think that's a better thing for me to write about, because no one else is going to cover it and it does present a different (and to me more interesting) view into how the tech is being used.

    In the meantime, when you own the site, you are then free to give me editorial direction. Until then, you're an anonymous troll, with poor reading comprehension ability.

  • On the post: My Kid Vibe Coded Their Way To Actually Learning Math

    Mike Masnick ( profile ), 27 May, 2026 @ 01:17pm

    I think we’d all prefer to see you put a little more focus on the downsides of AI tech and the industrialization associated with it, like data centers and the massive costs (economic and environmental) that those are having now and will have in the future.
    I've read a bunch about those things, but don't have a particularly strong opinion on them, and the focus of Techdirt for the 29 years it's been around is to write about things where I have a strong opinion about them, and don't feel they're being accurately portrayed elsewhere. There is plenty of discussion about data centers elsewhere. I honestly have zero idea what I could add to that discussion.
    And that’s to say nothing of the likes of deepfakes and AI-generated NCII⁠—like, say, 4chan’s /r/ board becoming so overrun with that shit that a single Wired story about the board becoming a haven for genAI NCII seems to have caused 4chan to take said board down.
    Interesting story, but not one that I was aware of. And, as per usual, at Techdirt I have a list of things I want to write about that is hundreds of stories long, and I get to maybe 10% of them. I honestly have little time for anyone who is like "how dare you not cover this story?" We cover what we want to cover and it's great to hear what else people would like us to cover, and we'll do as much as we can of what interests us (which again, I get to maybe 10% of what I can cover). If we had enough financial support to hire more writers to cover more, I would cover more. But, we have our very small crew and we cover what we find most interesting. That may or may not be what you find most interesting.

  • On the post: My Kid Vibe Coded Their Way To Actually Learning Math

    Mike Masnick ( profile ), 27 May, 2026 @ 10:11am

    So you're one of a few people here complaining that this is a bad take, but it's weird that no one ever explains why. This actually happened. I'm explaining what happened and why I found it interesting, while noting the also factual reality that the tools are a mess for many educational settings and causing real problems. Would you prefer that I just start lying about what the tech did to make you feel better? If so, then you should exit. I'm not going to lie to soothe your feelings about a technology. I will explain both its many limitations and problems as well as where it can be useful. If you would prefer being lied to, find another site to do that.

  • On the post: My Kid Vibe Coded Their Way To Actually Learning Math

    Mike Masnick ( profile ), 27 May, 2026 @ 10:09am

    It’s not that AI helped your child learn; it’s that AI was a tool that helped your child achieve a personal goal—writing this app—and learning happened along the way.
    Yes. I agree.

  • On the post: My Kid Vibe Coded Their Way To Actually Learning Math

    Mike Masnick ( profile ), 27 May, 2026 @ 10:08am

    Nothing in this post was "slopmongering." I detailed how the tools are being misused in many ways that damage education, but gave one example that was both interesting and surprising to me, but which I saw happen right here. Would you prefer I lie to you and pretend what happened didn't happen? If so... that's weird, and you're probably better served by sites that will lie to you to make you feel better.

  • On the post: My Kid Vibe Coded Their Way To Actually Learning Math

    Mike Masnick ( profile ), 27 May, 2026 @ 10:06am

    Ah. I think part of the issue here is thinking that ChatGPT is the right tool for this sort of thing. Earlier this year, I wrote about the many steps I took to seeing if I could get AI to write a Techdirt post, which also explains many of the steps I take in having it help me edit a post. And none of it is handing it off to a chatbot: https://www.techdirt.com/2026/02/18/how-close-can-ai-get-to-writing-a-techdirt-post/ The first thing is using agentic hooks, rather than a chatbot, which has a specific opinion. The second is then being able to create a very detailed (mine is many paragraphs long) system prompt so that the tool knows what it's doing. The third is then training it directly on a lot of my previous writing AS WELL AS a full style guide. And then, finally, not just providing it with what I wrote, but also all the source material as well. And then prompting it with a long and detailed prompt (not just "how would you fix this." The thing about these tools is that for them to work well in specific scenarios, they need context. And also the chatbots already have hidden system prompts that make them less useful for this sort of thing. And yes, my set up took a lot of work, but after being set up it's now simple (I actually write directly in the tool I built and click a button to do the review — and the tool makes suggested changes with red/greenlines in a sidebar which I can review/change/accept/reject.

  • On the post: My Kid Vibe Coded Their Way To Actually Learning Math

    Mike Masnick ( profile ), 26 May, 2026 @ 04:57pm

    It took a month and 5 days just to process a single article I wrote. The result was that it managed to introduce far more grammar mistakes than it would’ve fixed. It mangled several sentences and seemingly made changes for the sake of changes as well.
    I'm very curious what you did/what tools you used that this was the outcome.
    Using AI was awful and extremely tedious for me. I couldn’t ditch AI fast enough.
    Yeah. At no point am I suggesting it's right for every one or every situation. And I have seen plenty of people using it in a way that just makes more, not less, work for them. But that's... not really the point of this article?

  • On the post: My Kid Vibe Coded Their Way To Actually Learning Math

    Mike Masnick ( profile ), 26 May, 2026 @ 04:55pm

    12-year old, not 17. And, I'm a huge fan of Khan Academy (Sal Khan was my housemate around the time Techdirt was starting and it's always fun to follow what he's up to). For what it's worth he's also a big believer in using AI in education (though I think some of his predictions are misplaced): https://www.goodreads.com/en/book/show/196848520-brave-new-words But the simple fact is there is no one "right way" to learn, and different ways work for different people. Watching Khan Academy videos and using their tools is one way. Building tools for yourself is another.

  • On the post: ‘The Worst Leak I’ve Witnessed’: A CISA Contractor Left AWS GovCloud Credentials Sitting In A Public GitHub Repo

    Mike Masnick ( profile ), 26 May, 2026 @ 04:09pm

    Oh dude. Claiming that AFTER you fired or pushed out all the competent people THEN the remaining incompetence proves you were right all along that the agency was bad is... just next level stupidity.

  • On the post: Trump Sued Himself And Walked Away With A $100 Million Tax Debt Erased

    Mike Masnick ( profile ), 22 May, 2026 @ 12:44am

    It's made up nonsense. https://www.pbs.org/newshour/politics/ex-fbi-informant-charged-with-fabricating-biden-family-bribery-scheme-to-appear-in-court

  • On the post: The Science Is Not Settled: How Weak Evidence Is Fueling A National Push To Ban Social Media For Youth

    Mike Masnick ( profile ), 21 May, 2026 @ 03:36pm

    May have been a temporary glitch...

  • On the post: The Science Is Not Settled: How Weak Evidence Is Fueling A National Push To Ban Social Media For Youth

    Mike Masnick ( profile ), 21 May, 2026 @ 01:53pm

    Weird. It looks like when we did an update to WordPress it stopped autoposting to bsky. That's now fixed. We're not able to reproduce the contact Techdirt problem though (a test just worked). Do you have more details?

  • On the post: We Need A More Serious Discussion About Suicide And AI Chatbots

    Mike Masnick ( profile ), 19 May, 2026 @ 10:06am

    I am looking directly at the comment you replied to, and the comment I replied to before it. And they were clearly speaking of two specific claims: AI bros talking about jobs that would be destroyed and the claim that AI is going to make it harder for young people to have a future. There was no other specific claim about AI. And you insisted that I and Bluesky were supportive of such things, when we were not. As for "overselling and overpushing" the technology, again point to where? It's simply not true. Bluesky created a small, totally optional, side project staffed with a small team, to explore ways that AI could be useful for empowering users. It is not being pushed on anyone. The company isn't selling anything. And, as I've written, it's a useful exploration, but there is no desire by anyone to overhype what it can do or push it on anyone. You just keep lying. It's weird. And you should stop.

  • On the post: We Need A More Serious Discussion About Suicide And AI Chatbots

    Mike Masnick ( profile ), 18 May, 2026 @ 11:54pm

    If they’re responsible enough for that speech for it to fall under a first amendment protection, they’re responsible enough to be held liable for it.
    Truly, you are a work of genius. If you diagram out your sentence, it eventually reads: "if the speech qualifies not to be regulated, surely, it can be regulated." My goodness. Techdirt needs better trolls. I was hoping people here would have a discussion, not trot out ridiculous nonsensical platitudes.

  • On the post: We Need A More Serious Discussion About Suicide And AI Chatbots

    Mike Masnick ( profile ), 18 May, 2026 @ 11:51pm

    Oooooh. Look at that goalposts move. I asked you to name where Bluesky leadership had done the things in that post. You did not. Because you cannot. You are a liar. Anyway, I accept your effective admission that you lied in claiming Bluesky was saying things they never said. And, I also congratulate myself for accurately predicting you would not admit that you lied. You really should stop. But, I get it. You can't NOT lie about me. You're desperate to live in a fantasy world in your own head that doesn't match reality. I'm sorry that I refuse to join you in your fantasy world.

  • On the post: We Need A More Serious Discussion About Suicide And AI Chatbots

    Mike Masnick ( profile ), 18 May, 2026 @ 08:57pm

    Did you... did you... actually just use the fire in a crowded theater analogy on Techdirt?

  • On the post: We Need A More Serious Discussion About Suicide And AI Chatbots

    Mike Masnick ( profile ), 18 May, 2026 @ 08:56pm

    Can you point to where literally anyone in Bluesky's leadership team has suggested either "AI is going to drastically reduce the amount of available jobs" or "AI is going to make it harder for you to have a future"? I'll wait. Because you can't. If you were honest you would admit you were making shit up. But you won't. What the company has said is that if it can offer tools that put users in control over their own social experience, that's worth experimenting with. At no point is it forced on anyone. At no point is the company telling people they have to use it or insisting that it will take away jobs or anything of that nature. They're just experimenting with a totally optional tool whose guiding purpose is, as with everything Bluesky has built, putting the user in control. Look, I get it. It's fun to make up lies about what's going on. But it's pretty silly to just lie so consistently.

  • On the post: We Need A More Serious Discussion About Suicide And AI Chatbots

    Mike Masnick ( profile ), 18 May, 2026 @ 04:57pm

    Just to be clear (this came up in the last conversation on this topic, too), it's not the monkey selfie scenario at all. That was entirely about who can create a work that gets a copyright. And that is definitively "humans." That is not the same as "is this speech protected by the First Amendment?" There is plenty of speech that can get no copyright that is still protected by the First Amendment. Indeed, copyright itself is something of an exception to the First Amendment in the first place. So the two issues are entirely distinct.

Next >>