Along the vast divide between AI optimists and pessimists, the main factor in determining which side you fall on may be where you see yourself in the future: above the effects of AI, or beneath them.

[Jean] “Twenge is a psychology professor at San Diego State University and the author of “10 Rules for Raising Kids in a High-Tech World.” This op-ed article appeared in The New York Times, November 16, 2025.

The standardized test scores of American students had been rising for decades. Then they began to slide, dropping to their lowest point in two decades in 2023 and 2024.

This is not a problem confined to the United States. Worldwide, the performance of 15-year-olds in math, reading and science reached a nadir in 2022.

These dismal results are at least partly a result of the Covid-19 pandemic. Missed instruction during those years may still be having an impact on academic performance.

But that’s only part of the story. The decline in test scores started well before the pandemic, around 2012. One obvious culprit is smartphones, which became popular just as test scores started to decline. Since 2017, I’ve been doing research on what smartphones do to our mental health, and I recently started to study how they affect academic performance. The negative impact of smartphones on learning is one reason many school districts have instituted a bell-to-bell ban on smartphones in K-12 education, including all public schools in New York State, which also banned students’ personal laptops, tablets and smart watches.

That’s progress, especially when 83 percent of K-12 teachers surveyed by one major union think that smartphone bans are a good idea. But they are not a complete solution, because phones are not the only electronic devices students use at school. These days, nearly every middle and high school student — and a good number in the elementary grades as well — brings a laptop or tablet to school and uses it at home for homework.

Many of these devices are provided by schools. You might think that these school-issued devices allow only a limited number of functions, like access to classroom Canvas pages and Google Docs. If you assumed that, you would be wrong.

Sylvie McNamara, a parent of a ninth grader in Washington, D.C., wrote in Washingtonian magazine that her son was spending every class period watching TV shows and playing games on his school-issued laptop. He often had no idea what topics his classes were covering. When she asked school administrators to restrict her son’s use of the laptop, they resisted, saying the device was integral to the curriculum.

In a survey of American teenagers by the nonprofit Common Sense Media, one-fourth admitted they had seen pornographic content during the school day. Almost half of that group saw it on a school-issued device. Students watching porn in class doesn’t just affect the students themselves — picture being a teenager in math class trying to concentrate on sine and cosine while sitting behind that display of flesh. It is disturbing on a number of levels.

Even when laptop abuse doesn’t reach this point, it still consumes a substantial amount of instructional time. One study of Michigan State college students — nearly all legal adults presumably more capable of focusing their attention than young teens — found that they spent nearly 40 percent of class time scrolling social media, checking email or watching videos on their laptops — anything but their classwork.

School laptops are also distracting at home. Many allow unfettered access to YouTube, tempting students to watch an endless loop of videos instead of doing their homework. Just the other day, my daughter told me she was watching the violent cop show “The Rookie” on her school laptop at home. Apparently the device did not block access to the streaming service Disney+. How can we expect 13-year-olds to focus on their assignments when a vast library of entertaining video content is a tab away?

It seems ridiculous to have to say this, but digital distraction is terrible for academic performance. The more time college students spent doing something else on their laptops during class, the lower their exam scores, even after accounting for academic ability.

This also applies to teenagers around the world. A 2023 UNESCO report concluded that too much device use can hurt academic performance, mostly because of increased distraction and engagement in nonacademic activities.

In a study published in October in The Journal of Adolescence, I found that standardized test scores in math, reading and science fell significantly more in countries where students spent more time using electronic devices for leisure purposes during the school day than they did in countries where they spent less time.

The situation in Finland, once known for having one of the best school systems in the world, is telling. In 2022, teenagers in Finland admitted to using their devices during the school day for noneducational purposes for nearly 90 minutes. Perhaps as a result, the test scores of Finnish students plummeted between 2006 and 2022. In countries such as Japan, where students spend less than a half-hour on devices for leisure during the school day, academic performance has remained fairly steady, especially in math and science.

If tablets and laptops are behind even a small portion of the decline in academic performance, parents and educators will need to work together to find solutions. At the moment, parents are virtually powerless: They can’t install parental control software on school devices. Nevertheless, many districts try to foist responsibility back onto parents by telling them, as my children’s district does, that “there is no substitute for parental supervision. Be knowledgeable of what sites your child goes to online.” How, exactly, are we supposed to do that when we can’t install control software and given that it’s not possible to hover over our teenagers every minute?

If school districts want to improve their test scores — and most are desperate to do so — changing the way students use school-issued devices is critical. To start, school I.T. departments should lock down devices much more securely so students can’t use them to watch TV shows, play games or continuously consume videos. Whatever efforts schools are making in that direction right now are frequently evaded by tech-savvy students.

There should be districtwide policies that specifically disallow these types of uses and instruct teachers to embed educational videos on the classroom page instead of giving students unlimited access to YouTube.

Districts and teachers should also consider scaling back on the number of assignments that require a device to complete in the first place. A paper math worksheet or a handwritten response to a reading assignment is one less opportunity for kids to use a device chock-full of digital distractions, and one less opportunity to cut and paste an essay written by ChatGPT.

Parents should also have the option to opt out. I’ve spoken with many parents whose children struggle to focus while using laptops, only for school administrators to tell them the devices are required. Angela Arsenault, a state representative in Vermont, is planning to introduce a bill to give parents the ability to opt their children out of receiving school-issued devices. A version of the bill was first introduced in 2015, a stark illustration of how long this has been an issue.

Districts could even eliminate school electronic devices entirely. Many parents and teachers might protest that this would have an adverse impact on learning or tilt the scales in favor of wealthier students who have access to their own devices, but several studies suggest it might instead improve learning. One study of nearly 300,000 fourth and eighth graders in the United States found that students who spent more time using digital devices in language arts classes performed worse on reading tests. A 2018 meta-analysis found that reading on paper, compared with reading digitally, led to significantly better comprehension among students, from elementary school to college. Across 24 studies, college students who took handwritten notes were 58 percent more likely to get A’s in their courses than those who typed notes on laptops. In contrast, students who typed notes were 75 percent more likely to fail the course than those who wrote them by hand.

Although it once seemed like a good idea to give every child his or her own device, it’s clear that those policies have been a failure. It may be possible to harness the power of school devices more judiciously, with little to no device use in lower grades, and high school students given laptops strictly limited to relevant apps. We could go further, creating completely device-free schools with rare exceptions for students with special needs. It would be back to the textbooks, paper and pencil of previous eras — when the most significant classroom distraction was students passing notes.

Many adults struggle to concentrate on work when social media, shopping and movies are just a click away. Imagine how much more difficult it is for a 16-year-old, much less an 11-year-old, to focus in the same situation. Asking students to drill down on their schoolwork amid an array of digital distractions isn’t just bad for test scores; it is inimical to learning.

And it is fundamentally unfair to our children.

U.S. adults under 30 follow news less closely than any other age group. And they’re more likely to get (and trust) news from social media.

The post Young Adults and the Future of News appeared first on Pew Research Center.

In mathematics, ubiquitous objects called groups display nearly magical powers. Though they’re defined by just a few rules, groups help illuminate an astonishing range of mysteries. They can tell you which polynomial equations are solvable, for instance, or how atoms are arranged in a crystal. And yet, among all the different kinds of groups, one type stands out. Identified in the early 1870s…

Source

I observe and interact with humans between the ages of eighteen and twenty-five almost every day of my life, as part of my job. Like many, when AI chatbots were maliciously (indiscriminately is too kind of a word) released into the wild several years ago, I watched in horror—the kind of horror one might feel watching a pod of orcas play with a baby seal before ripping it to shreds. Only in this case the seal believes itself to be the one having all the fun. That young people were attracted to an openly available technological wonder is not surprising, though it was disconcerting to witness. What was truly shocking to me in that first year—and still is—was the almost complete lack of resistance by those charged with educating these young souls. Half-hearted appeals for deliberate discernment of the adoption of AI into higher education quickly dissipated and joined the chorus of administrators recommending “responsible use,” by this very phrase unintentionally treating AI as a harmful substance, yet one we can’t expect young people to abstain from using.

My own view from the beginning of this new era of machine domination has been that tools of such power require their users to have an extremely high degree of maturity and responsibility. But the whole growth model of corporate AI tools requires as much input and interaction as possible from as many users as possible. Thus, it came to be that perhaps the most powerful tool ever created by humanity fell into the untried hands of a generation known for, among other things, eating Tide pods. Not that all the blame for the irresponsible use of AI falls on younger generations. Aspirations to transgress the bounds of natural limitations have plagued humanity from the beginning. “Come, let us build for ourselves a city, and a tower with its top in the heavens, and let us make a name for ourselves,” Babel’s vision-casting team said. The builders and sellers of AI—futurists, pioneers, entrepreneurs, and engineers, along with their funders—will be the most culpable for harm caused, especially if products are designed to entice and addict.

In the many conversations I have had on this topic over the past three years, most of my interlocutors have taken either a neutral or a positive stance toward AI, though of late I have seen some who were previously unconcerned begin to grow wary. I find that the majority of those I speak with on the topic, however, have given little thought to the potential downsides of AI. Few have made the effort to subject their use of AI to moral scrutiny, and even fewer have dared to ask whether any and all use of AI may implicate the user in some moral evil. Yet there are a multitude of reasons not merely to approach AI with caution but to engage in determined opposition to it. In what follows I offer a handful of what I take as the weightier among these reasons, and I invite those who read this to consider joining the resistance.

The most common reactions I hear in response to my rejection of AI have to do with all the real or imagined goods AI might help bring about: medical advancement, all manner of scientific research, eliminating dull and time-consuming tasks in human work, and so on. Virtually everyone acknowledges the risks of AI and the ways it can be used for harm rather than good, but these are taken as the price that must be paid to attain the good. AI is simply a tool, they say, and like all tools it can be used for good or for ill. It would be ludicrous to get rid of the tool just because some bad actors abuse it. This line of thought fails in at least two ways.

First of all, it fails to recognize a distinction between moral evil and natural evil. Natural evils are things we deem bad—usually due to the loss of life—but for which we are not culpable: natural disasters, diseases, etc. Moral evils are things we deem bad and hold people responsible for. Moral evils are always more grave for the souls involved than natural evils, even if the scale seems smaller. AI may help attenuate natural evils by finding cures for diseases or drastically improving disaster prediction, for example. But the capability and availability of AI have already drastically increased moral evil. This is not a net gain for humanity. To put it bluntly, it is better to suffer natural evils and possess virtue than to eradicate natural evils and lack virtue. Ends never justify means.

Second, although AI can be thought of as a tool, it is unlike any other tool due to its ability to “self-improve” based on human input and interaction along with ever-expanding data sets. If a person uses a hammer to smash the skull of an innocent other, this is clearly a case of a tool being misused. But the hammer undergoes no change; through the act of someone using it to commit murder, the hammer itself does not become a better weapon. With AI this is not the case. As more and more people use AI tools, the tools themselves change and get better at doing what their users ask. Which means even supposedly harmless uses of AI actively contribute to the tools becoming better at bringing ill will to fruition. Millions of users, for example, creating relatively innocuous fake videos for work or for fun contribute to the tool becoming better at making realistic child pornography. Again, this is not a tradeoff we should be willing to tolerate.

In addition to these moral questions pertaining to the very existence and use of AI, there are moral concerns pertaining to the peripheries of the AI complex. A prime example of this is the investments being poured into creating an AI-integrated world. Hundreds of billions of dollars have already been spent on developing and deploying AI tools, and there are only signs that this will increase in the near future. If achieving an AI-integrated world were a matter of meeting critical needs, perhaps such expenditures would be justified. But there are no critical needs in the world that can only be addressed with AI. Poverty, violence, famine, loneliness, lack of healthcare, environmental collapse—these are needs which, in order to be met, require human and financial resources, not AI. Because these critical needs have not been met, it is unjust to direct such vast resources toward something which, though technologically revolutionary, is quite frivolous in relation to the realities of human existence. We do not need AI to flourish as humans. We do need peace, healthy communities, clean water, fertile soil, and a dependable food supply, all of which AI cannot supply.

The environmental cost of AI is something to consider as well, working on the assumption that keeping this planet perpetually fruitful and safe to inhabit is important for humanity. Massive data centers are required to keep AI running, along with cloud computing, streaming, and a host of other digital commodities, and each data center requires a massive amount of electrical and water supply, not to mention actual land usage. Data centers collectively are becoming one of the leading emitters of carbon through electrical consumption, and newer “hyperscale” centers are expected to require more water than current ones. On a planet that was already facing environmental crises before the rise of AI, surely AI’s energy, water, mined material, and land demands are going to make it even more difficult to address these crises, especially since AI seems to have quickly been deemed a necessity by corporate and political powers. Care for our common home ought to at least cause us to question whether this is the best path forward for humanity.

There are a host of other reasons to reject the use of AI outright, such as the minor issue of a possible machine-assisted human extinction event, but I will offer one final thought. The use of AI degrades the value of human work and so of humans themselves. Those who tout the benefits of AI, like most advocates of machine technology, view natural human limitations as weaknesses to be transcended if possible. This attitude is of course a derivative of an economic order built on hyper consumption. In both production and consumption, more and faster is always better. Since the introduction of machines greatly enhances production, the goodness of machines is not to be questioned. But the unmentioned corollary of this is that the productive value of a mere human without a machine is reduced to almost nothing. Being human is not enough. You and I, considered in ourselves, are of insufficient value in the minds of those pushing for an AI-integrated world, unless we amend our humanity with machines.

The effect of this in the agricultural realm is widely recognized. Even though the only things required to make land productive are human care, human or animal power, and water, a farm without machines is a near impossibility today. And the idea of farm work being done without machines is seen as something rightly relegated to some past peasantry. This same degradation of work done at an actual human scale by actual humans will come to every sector that adopts AI. Productivity will go up, value and diversity will go down, and the monoculture dullness that one sees flying over the machine-leveled fields of middle America, with its ecological and cultural costs, will characterize much of society and the economy. Yes, you will be able to escape from this ugliness into virtual worlds. But I, for one, would prefer to enjoy the beauty of the real world—the beauty that is the ordered harmony of great variety; the beauty that must be cultivated, preserved, and fought for; the beauty that can only exist when natural limits are respected; the beauty that came under threat after the industrial revolution; the beauty that could very well disappear after the AI revolution. Unless we resist.

Image Credit.

I don't like RSS readers. I know, this is blasphemous especially on a website where I'm actively encouraging you to subscribe through RSS. As someone writing stuff, RSS is great for me. I don't have to think about it, the requests are pretty light weight, I don't need to think about your personal data or what client you are using. So as a protocol RSS is great, no notes.

However as something I'm going to consume, it's frankly a giant chore. I feel pressured by RSS readers, where there is this endlessly growing backlog of things I haven't read. I rarely want to read all of a websites content from beginning to end, instead I like to jump between them. I also don't really care if the content is chronological, like an old post about something interesting isn't less compelling to me than a newer post.

What I want, as a user experience, is something akin to TikTok. The whole appeal of TikTok, for those who haven't wasted hours of their lives on it, is that I get served content based on an algorithm that determines what I might think is useful or fun. However what I would like is to go through content from random small websites. I want to sit somewhere and passively consume random small creators content, then upvote some of that content and the service should show that more often to other users. That's it. No advertising, no collecting tons of user data about me, just a very simple "I have 15 minutes to kill before the next meeting, show me some random stuff."

In this case the "algorithm" is pretty simple: if more people like a thing, more people see it. But with Google on its way to replacing search results with LLM generated content, I just wanted to have something that let me play around with the small web the way that I used to.

There actually used to be a service like this called StumbleUpon which was more focused on pushing users towards popular sites. It has been taken down, presumably because there was no money in a browser plugin that sent users to other websites whose advertising you didn't control.

TL;DR

You can go download the Firefox extension now and try this out and skip the rest of this if you want. https://timewasterpro.xyz/ If you hate it or find problems, let me know on Mastodon. https://c.im/@matdevdug

Functionality

So I wanted to do something pretty basic. You hit a button, get served a new website. If you like the website, upvote it, otherwise downvote it. If you think it has objectionable content then hit report. You have to make an account (because I couldn't think of another way to do it) and then if you submit links and other people like it, you climb a Leaderboard.

On the backend I want to (very slowly so I don't cost anyone a bunch of money) crawl a bunch of RSS feeds, stick the pages in a database and then serve them up to users. Then I want to track what sites get upvotes and return those more often to other users so that "high quality" content shows up more often. "High quality" would be defined by the community or just me if I'm the only user.

It's pretty basic stuff, most of it copied from tutorials scattered around the Internet. However I really want to drive home to users that this is not a Serious Thing. I'm not a company, this isn't a new social media network, there are no plans to "grow" this concept beyond the original idea unless people smarter than me ping with me ideas. So I found this amazing CSS library: https://sakofchit.github.io/system.css/

The Apple's System OS design from the late-80s to the early 90s was one of my personal favorites and I think would send a strong signal to a user that this is not a professional, modern service.

Great, the basic layout works. Let's move on!

Backend

So I ended up doing FastAPI because it's very easy to write. I didn't want to spend a ton of time writing the API because I doubt I nailed the API design on the first round. I use sqlalchemy for the database. The basic API layout is as follows:

• admin - mostly just generating read-only reports of like "how many websites are there"
• leaderboard - So this is my first attempt at trying to get users involved. Submit a website that other people like? Get points, climb leaderboard.

The source for the RSS feeds came from the (very cool) Kagi small web Github. https://github.com/kagisearch/smallweb. Basically I assume that websites that have submitted their RSS feeds here are cool with me (very rarely) checking for new posts and adding them to my database. If you want the same thing as this does, but as an iFrame, that's the Kagi small web service.

The scraping work is straightforward. We make a background worker, they grab 5 feeds every 600 seconds, they check for new content on each feed and then wait until the 600 seconds has elapsed to grab 5 more from the smallweb list of RSS feeds. Since we have a lot of feeds, this ends up look like we're checking for new content less than once a day which is the interval that I want.

Then we write it out to a sqlite database and basically track "has this URL been reported", if so, put it into a review queue and then how many times this URL has been liked or disliked. I considered a "real" database but honestly sqlite is getting more and more scalable every day and its impossible to beat the immediate start up and functionality. Plus very easy to back up to encrypted object storage which is super nice for a hobby project where you might wipe the prod database at any moment.

In terms of user onboarding I ended up doing the "make an account with an email, I send a link to verify the email". I actually hate this flow and I don't really want to know a users email. I never need to contact you and there's not a lot associated with your account, which makes this especially silly. I have a ton of email addresses and no real "purpose" in having them. I'd switch to Login with Apple, which is great from a security perspective but not everybody has an Apple ID.

I also did a passkey version, which worked fine but the OSS passkey handling was pretty rough still and most people seem to be using a commercial service that handled the "do you have the passkey? Great, if not, fall back to email" flow. I don't really want to do a big commercial login service for a hobby application.

Auth is a JWT, which actually was a pain and I regret doing it. I don't know why I keep reaching for JWTs, they're a bad user experience and I should stop.

Can I just have the source code?

I'm more than happy to release the source code once I feel like the product is in a somewhat stable shape. I'm still ripping down and rewriting relatively large chunks of it as I find weird behavior I don't like or just decide to do things a different way.

In the end it does seem to do whats on the label. We have over 600,000 individual pages indexed.

So how is it to use?

Honestly I've been pretty pleased. But there are some problems.

First I couldn't find a reliable way of switching the keyboard shortcuts to be Mac/Windows specific. I found some options for querying platform but they didn't seem to work, so I ended up just hardcoding them as Alt which is not great.

The other issue is that when you are making an extension, you spend a long time working with these manifests.json. The specific part I really wasn't sure about was:

"browser_specific_settings": {
    "gecko": {
      "id": "admin@timewasterpro.xyz",
      "strict_min_version": "80.0",
      "data_collection_permissions": {
        "required": ["authenticationInfo"]
      }
    }
  }

I'm not entirely sure if that's all I'm doing? I think so from reading the docs.

Anyway I built this mostly for me. I have no idea if anybody else will enjoy it. But if you are bored I encourage you to give it a try. It should be pretty light weight and straight-forward if you crack open the extension and look at it. I'm not loading any analytics into the extension so basically until people complain about it, I don't really know if its going well or not.

Future stuff

• I need to sort stuff into categories so that you get more stuff in genres you like. I don't 100% know how to do that, maybe there is a way to scan a website to determine the "types" of content that is on there with machine learning? I'm still looking into it.
• There's a lot of junk in there. I think if we reach a certain number of downvotes I might put it into a special "queue".
• I want to ensure new users see the "best stuff" early on but there isn't enough data to determine "best vs worst".
• I wish there were more independent photography and science websites. Also more crafts. That's not really a "future thing", just me putting a hope out into the universe. Non-technical beta testers get overwhelmed by technical content.