Nuclear Monster

  • I’ve been complaining about how Duolingo punishes users for mistakes for a long time now, they’re now moving from hearts that you lose when you make a mistake, to an energy mechanic according to this article from Jay Peters writing for The Verge:

    “Under the new system, you’ll spend one unit of energy to complete an exercise, and a mistake will cost one energy. But you’ll also get extra energy back at a randomized rate for completing multiple lessons right in a row. As a result, users should be able to do more lessons, and that’s what’s showing up in the data, Moses Wayne, a senior staff engineer at Duolingo, tells The Verge.“

    There’s a bunch of math that goes into this to figure out that it will be similarly punishing for free users as the current Heart-based system. As before, Duolingo subscribers get unlimited Energy.

    I’ve also noticed that Duolingo’ve added a Chess course, which I’ve only tried a little bit.

    It feels like the Duolingo company is in Emergency save-our-reputation mode with these changes. I don’t believe the change to Energy from Hearts is good enough, it still slows down free users from focusing on learning with whatever time they have available and will likely end-up penalizing people for making mistakes. They should either move to subscription only or just have more ads for free users instead of slowing them down. What is the cost to the company for a free user these days? It can’t be much.

    Here’s my sure-fire way to save the company’s reputation:

    • Take the company private
    • Drop all internal AI slop and plans for the slop
    • Hire the workers back
    • Make the company a worker-owned co-op
    • Luis von Ahn retires.
    • Users now fully on-board, move to subscriptions only for most users, let users buy each other subscriptions and offer no-questions-asked discounts for anyone who can’t afford the full-price subscription.

  • Yesterday Duolingo’s CEO Luis Von Ahn re-confirmed that the slop is yummy in a post to Linkedin that was also sent to the entire company:

    “I’ve said this in Q&As and many meetings, but I want to make it official: Duolingo is going to be AI-first.

    AI is already changing how work gets done. It’s not a question of if or when. It’s happening now. When there’s a shift this big, the worst thing you can do is wait. In 2012, we bet on mobile. While others were focused on mobile companion apps for websites, we decided to build mobile-first because we saw it was the future. That decision helped us win the 2013 iPhone App of the Year and unlocked the organic word-of-mouth growth that followed. Betting on mobile made all the difference. We’re making a similar call now, and this time the platform shift is AI.”

    Von Ahn goes on to confirm that they won’t hire new workers without teams first proving that AI can’t do the job, contractors will be laid off and replaced with AI when possible and they’re already generating content with AI slop machines:

    “AI isn’t just a productivity boost. It helps us get closer to our mission. To teach well, we need to create a massive amount of content, and doing that manually doesn’t scale. One of the best decisions we made recently was replacing a slow, manual content creation process with one powered by AI. Without AI, it would take us decades to scale our content to more learners. We owe it to our learners to get them this content ASAP.

    AI also helps us build features like Video Call that were impossible to build before. For the first time ever, teaching as well as the best human tutors is within our reach.

    Being AI-first means we will need to rethink much of how we work. Making minor tweaks to systems designed for humans won’t get us there. In many cases, we’ll need to start from scratch. We’re not going to rebuild everything overnight, and some things like getting AI to understand our codebase-will take time. However, we can’t wait until the technology is 100% perfect. We’d rather move with urgency and take occasional small hits on quality than move slowly and miss the moment.“

    We’ll be rolling out a few constructive constraints to help guide this shift:

    • We’ll gradually stop using contractors to do work that AI can handle
    • AI use will be part of what we look for in hiring
    • AI use will be part of what we evaluate in performance reviews
    • Headcount will only be given if a team cannot automate more of their work
    • Most functions will have specific initiatives to fundamentally change how they work

    All of this said, Duolingo will remain a company that cares deeply about its employees. This isn’t about replacing Duos with AI. It’s about removing bottlenecks so we can do more with the outstanding Duos we already have. We want you to focus on creative work and real problems, not repetitive tasks. We’re going to support you with more training, mentorship, and tooling for AI in your function.

    Change can be scary, but I’m confident this will be a great step for Duolingo. It will help us better deliver on our mission and for Duos, it means staying ahead of the curve in using this technology to get things done.

    –Luis”

    I’ve written that I’m increasingly unhappy with the service for getting away from their original goals of “free. fun. effective” in order to focus on more AI slopification.

    Duolingo is a publicly traded company and it might be unacceptable to their shareholders to not ruin the company with AI slop, but there has to be a way they could do it without laying off contractors and ruining their service. I also do not care about their shareholders and believe it was a mistake to make the company publicly traded in the first place. This decision is an embarrassment to Duolingo and terrible for learning.

    I believe this is the time to focus more on other tools that lean more towards non-slop. There are a lot of them on my language learning resources page.

    There’s this old Twitter post I love that I think perfectly describes why anyone is chasing this “opportunity”:

    “Being a billionaire must be insane. You can buy new teeth, new skin. All your chairs cost 20,000 dollars and weigh 2,000 pounds. Your life is just a series of your own preferences. In terms of cognitive impairment it’s probably like being kicked in the head by a horse every day”

  • Cherry claims to be the oldest company making keyboards. They invented the MX keyboard switch standard that the majority of mechanical keyboard switches are based on, and have been producing computer keyboards in Germany since 1973.  Three days ago Cherry announced they’re ending production in Germany and moving it to China:

    “The restructuring concept includes, among other measures, the complete discontinuation of switch production at the Auerbach site and its transfer to a partner in China. In this context, jobs will also be reduced in a socially responsible manner in consultation with the works council. The Auerbach site will be retained and is to be transformed into a development, logistics, and service hub for Europe.”

    Absolute tragedy for those workers who will lose their jobs there and the end of a legendary production line.

  • Panic announced today a new set of 12 games coming in the second season of them for their little yellow handheld console called the Playdate.

    If you remember, when the Playdate launched in 2022 it included 24 games in what Panic called Season One. With that season you could either get two games from the 24 every week or have them all available at once. Anyone who got the handheld afterwards could make the same choice. Panic wants to do something similar for this second season of 12 games.

    We only get a little bit of information about the new season of games, but I am really curious about Fulcrum Defender from Subset Games who also developed FTL and Into The Breach. Fulcrum Defender looks like an arcady good time.

    Panic also hinted at more from whatever Blippo+ is during the announcement video for Playdate Season 2.

  • This is a post I’ve been meaning to write for a few years, as I’ve grown more dissatisfied with Duolingo. Then a friend posted that they ran out of hearts during a lesson, the only way forward was to subscribe, causing them to lose their hard-earned streak, and they’re quitting.

    I’ve been using Duolingo to learn German on and off for over a decade now. A few years ago when the pandemic started* I made Duolingo part of a routine that I’ve been streaming most days on Twitch. My streak is now at 1768 days.

    When Duolingo first started over a decade ago, it was a language learning business with a focus on free education and courses made by people. Even today the three most important words on the Duolingo homepage when you’re logged out are “free. fun. effective” This mantra is repeated a few times on the page.

    Over time the name Duolingo has become synonymous with language learning, but I believe it has now gone too far into a focus on profits, and away from their original goals. I’ll spell out a few reasons why.

    Duolingo’s entire game and marketing focus on streaks of daily learning can help form a good habit, but losing a streak for unavoidable reasons is why some people stop learning. People have to be encouraged to not let losing the streak end learning.

    Learning doesn’t happen without mistakes and Duolingo now makes it more difficult to learn without a subscription. Duolingo has long had “hearts” that you lose like a video game character’s health when you make a mistake. The change now is that users can’t earn more hearts by practicing their target language.

    When you make a mistake and run out of hearts, Duolingo makes this experience frustrating and robs users of fun by forcing them to give up progress on a lesson and start over, or subscribe to Duolingo for unlimited hearts.

    I believe this is the worst part of the modern Duolingo experience for free users, but there are plenty of other annoying details of the modern version of Duolingo.

    Even if you have been a Super Duolingo subscriber, they are now pushing Plus subscribers to use modern AI slop at another, more expensive subscription tier, Duolingo Max. I subscribe to the Super Duolingo family plan for $120 a year which is a lot for us, but somehow isn’t enough for Duolingo. The company has become extremely pushy to get subscribers to upgrade to their $240 Duolingo Max subscription for families. Buttons are being replaced on the primary menu for Duolingo to buttons that have no function for Super Duolingo users, they just launch into ads for Duolingo Max.

    Duolingo got rid of their built-in forums for each lesson that let users help other users ask questions and understand tricky bits of their target language. One of the features of Duolingo Max is an AI slop explanation for why your answer to a Duolingo lesson was wrong. Here the Duolingo company has exchanged real answers from real people for slop answers that are helping to incinerate our planet through thermally inefficient computing resources at a higher cost.

    The web version of Duolingo gives users far fewer experience points while still pushing people to compete on the same leaderboards with iOS and Android Duolingo app users who receive far more experience points for the same lessons. When I look at my next lesson on the Duolingo website, it is offering me 10 experience points, the same lesson on the mobile app is offering 35.

    At the start of 2024 Duolingo fired about 10% of their workforce in order to focus on AI. I’ve seen reports that they’ve also cut their support team down to just one full-time worker though they may have some contractors supporting that role.

    There are good things that Duolingo has done. The mostly adorable characters that act out and speak in the lessons are extremely helpful for some types of lessons. Duolingo recently added music and math lessons that could be very valuable. They haven’t started to crack down yet on people sharing their family accounts with others outside of their households like so many other companies have. However, I have moved more of my language learning to other tools and resources.

    Especially for learning German, I’ve found that I am more effective at learning when I hear real voices instead of text-to-speech generated voices and especially when I can see actual human faces speaking the words I’m having the most difficulty pronouncing.

    For all of these reasons, but especially because I see Duolingo having diverged so much from their core values of “free. fun. effective.” at this point I don’t value Duolingo as much in 2025 as I now value tools like Seedlang that use real people in video lessons giving me a more effective experience in learning German. When I meet people who are interested in learning a language I’m also much less likely to recommend Duolingo to them because I know how easy it is to lose a streak for free users, get disappointed, and give up learning entirely just like my friend’s post that caused me to write this.

    It is extremely unfortunate that Duolingo have squandered their success. I hope that the company changes, there is still a great deal of value in a gamified learning experience with cute characters that rewards learning at your own pace. They could even eliminate the free tier and just require a subscription after a certain amount of progress in a course. In my opinion that would at least be a more honest experience than frustrating users with broken streaks and lost hearts. What kind of a teacher puts these roadblocks in the path of learning? Not a very good one.

    *Did you know that tens of thousands of people are still dying of COVID every year? It’d be good if you continued to mask up and avoid spreading this disease.