Cal Newport's Blog, page 3

AI and Work (Some Predictions)

One of the main topics of this newsletter is the quest to cultivate sustainable and meaningful work in a digital age. Given this objective, it’s hard to avoid confronting the furiously disruptive potentials of AI.

I’ve been spending a lot time in recent years, in my roles as a digital theorist and technology journalist, researching and writing about this topic, so it occurred to me that it might be useful to capture in one place all of my current thoughts about the intersection of AI and work.

The obvious caveat applies: these predictions will shift — perhaps even substantially — as this inherently unpredictable sector continues to evolve. But here’s my current best stab at what’s going on now, what’s coming soon, and what’s likely just hype.

Let’s get to it…

Where AI Is Already Making a Splash

When generative AI made its show-stopping debut a few years ago, the smart money was on text production becoming the first killer app. For example, business users, it was thought, would soon outsource much of the tedious communication that makes up their day — meeting summaries, email, reports — into AI tools.

A fair amount of this is happening, especially when it comes to lengthy utilitarian communication where the quality doesn’t matter much. I recently attended a men’s retreat, for example, and it was clear that the organizer had used ChatGPT to create the final email summarizing the weekend schedule. And why not? It got the job done and saved some time.

It’s becoming increasingly clear, however, that for most people the act of writing in their daily lives isn’t a major problem that needs to be solved, which is capping the predicted ubiquity of this use case. (A survey of internet users found that only around 5.4% had used ChatGPT to help write emails and letters. And this includes the many who maybe experimented with this capability once or twice before moving on.)

The application that has instead leaped ahead to become the most exciting and popular use of these tools is smart search. If you have a question, instead of turning to Google you can query a new version of ChatGPT or Claude. These models can search the web to gather information, but unlike a traditional search engine, they can also process the information they find and summarize for you only what you care about. Want the information presented in a particular format, like a spreadsheet or a chart? A high-end model like GPT-4o can do this for you as well, saving even more extra steps.

Smart search has become the first killer app of the generative AI era because, like any good killer app, it takes an activity most people already do all the time — typing search queries into web sites — and provides a substantially, almost magically better experience. This feels similar to electronic spreadsheets conquering paper ledger books or email immediately replacing voice mail and fax. I would estimate that around 90% of the examples I see online right now from people exclaiming over the potential of AI are people conducting smart searches.

This behavioral shift is appearing in the data. A recent survey conducted by Future found that 27% of US-based respondents had used AI tools such as ChatGPT instead of a traditional search engine. From an economic perspective, this shift matters. Earlier this month, the stock price for Alphabet, the parent company for Google, fell after an Apple executive revealed that Google searches through the Safari web browser had decreased over the previous two months, likely due to the increased use of AI tools.

Keep in mind, web search is a massive business, with Google earning over $175 billion from search ads in 2023 alone. In my opinion, becoming the new Google Search is likely the best bet for a company like OpenAI to achieve profitability, even if it’s not as sexy as creating AGI or automating all of knowledge work (more on these applications later).

The other major success story for generative AI at the moment is computer programming. Individuals with only rudimentary knowledge of programming languages can now produce usable prototypes of simple applications using tools like ChatGPT, and somewhat more advanced projects with AI-enhanced agent-style helpers like Roo Code. This can be really useful for quickly creating tools for personal use or seeking to create a proof-of-concept for a future product. The tech incubator Y Combinator, for example, made waves when they reported that a quarter of the start-ups in their Winter 2025 batch generated 95% or more of their product’s codebases using AI.

How far can this automated coding take us? An academic computer scientist named Judah Diament recently went viral for noting that the ability for novice users to create simple applications isn’t new. There have been systems dedicated to this purpose for over four decades, from HyperCard to VisualBasic to Flash. As he elaborates: “And, of course, they all broke down when anything slightly complicated or unusual needs to be done (as required by every real, financially viable software product or service).”

This observation created major backlash — as does most expressions of AI skepticism these days — but Diament isn’t wrong. Despite recent hyperbolic statements by tech leaders, many professional programmers aren’t particularly worried that their jobs can be replicated by language model queries, as so much of what they do is experience-based architecture design and debugging, which are unrelated skills for which we currently have no viable AI solution.

Software developers do, however, use AI heavily: not to produce their code from scratch, but instead as helper utilities. Tools like GitHub’s Copilot are integrated directly into the environments in which these developers already work, and make it much simpler to look up obscure library or AI calls, or spit out tedious boilerplate code. The productivity gains here are notable. Programming without help from AI is rapidly becoming increasingly rare.

The Next Big AI Application

Language model-based AI systems can respond to prompts in pretty amazing ways. But if we focus only on outputs, we underestimate another major source of these models’ value: their ability to understand human language. This so-called natural language processing ability is poised to transform how we use software.

There is a push at the moment, for example, led by Microsoft and its Copilot product (not to be confused with GitHub Copilot), to use AI models to provide natural language interfaces to popular software. Instead of learning complicated sequences of clicks and settings to accomplish a task in these programs, you’ll be able to simply ask for what you need; e.g., “Hey Copilot, can you remove all rows from this spreadsheet where the dollar amount in column C is less than $10 dollars then sort everything that remains by the names in Column A? Also, the font is too small, make it somewhat larger.”

Enabling novice users to access to expert-level features in existing software will aggregate into huge productivity gains. As a bonus, the models required to understand these commands don’t have to be nearly as massive and complicated as the current cutting-edge models that the big AI companies use to show off their technology. Indeed, they might be small enough to run locally on devices, making them vastly cheaper and more efficient to operate.

Don’t sleep on this use case. Like smart search, it’s also not as sexy as AGI or full automation, but I’m increasingly convinced that within the next half-decade or so, informally-articulated commands are going to emerge as one of the dominate interfaces to the world of computation.

What About Agents?

One of the more attention-catching storylines surrounding AI at the moment is the imminent arrival of so-called agents which will automate more and more of our daily work, especially in the knowledge sectors once believed to be immune from machine encroachment.

Recent reports imply that agents are a major part of OpenAI’s revenue strategy for the near future. The company imagines business customers paying up to $20,000 a month for access to specialized bots that can perform key professional tasks. It’s the projection of this trend that led Elon Musk to recently quip: “If you want to do a job that’s kinda like a hobby, you can do a job. But otherwise, AI and the robots will provide any goods and services that you want.”

But progress in creating these agents has recently slowed. To understand why requires a brief snapshot of the current state of generative AI technology…

Not long ago, there was a belief in so-called scaling laws that argued, roughly speaking, that as you continued to increase the size of language models, their abilities would continue to rapidly increase.

For a while this proved true: GPT-2 was much better than the original GPT, GPT-3 was much better than GPT-2, and GPT-4 was a big improvement on GPT-3. The hope was that by continuing to scale these models, you’d eventually get to a system so smart and capable that it would achieve something like AGI, and could be used as the foundation for software agents to automate basically any conceivable task.

More recently, however, these scaling laws have begun to falter. Companies continue to invest massive amounts of capital in building bigger models, trained on ever-more GPUs crunching ever-larger data sets, but the performance of these models stopped leaping forward as much as they had in the past. This is why the long-anticipated GPT-5 has not yet been released, and why, just last week, Meta announced they were delaying the release of their newest, biggest model, as its capabilities were deemed insufficiently better than its predecessor.

In response to the collapse of the scaling laws, the industry has increasingly turned its attention in another direction: tuning existing models using reinforcement learning.

Say, for example, you want to make a model that is particularly good at math. You pay a bunch of math PhDs $100 an hour to come up with a lot of math problems with step-by-step solutions. You then take an existing model, like GPT-4, and feed it these problems one-by-one, using reinforcement learning techniques to tell it exactly where it’s getting certain steps in its answers right or wrong. Over time, this tuned model will get better at solving this specific type of problem.

This technique is why OpenAI is now releasing multiple, confusingly-named models, each seemingly optimized for different specialties. These are the result of distinct tunings. They would have preferred, of course, to simply produce a GPT-5 model that could do well on all of these tasks, but that hasn’t worked out as they hoped.

This tuning approach will continue to develop interesting tools, but it will be much more piecemeal and hit-or-miss than what was anticipated when we still believed in scaling laws. Part of the difficulty is that this approach depends on finding the right data for each task you want to tackle. Certain problems, like math, computer programming, and logical reasoning, are well-suited for tuning as they can be described by pairs of prompts and correct answers. But this is not the case for many other business activities, which can be esoteric and bespoke to a given context. This means many useful activities will remain un-automatable by language model agents into the foreseeable future.

I once said that the real Turing Test for our current age is an AI system that can successfully empty my email inbox, a goal that requires the mastery of any number of complicated tasks. Unfortunately for all of us, this is not a test we’re poised to see passed any time soon.

Are AGI and Superintelligence Imminent?

The Free Press recently published an article titled “AI Will Change What it Means to Be Human. Are We Ready?”. It summarized a common sentiment that has been feverishly promoted by Silicon Valley in recent years: that AI is on the cusp of changing everything in unfathomably disruptive ways.

As the article argues:

OpenAI CEO Sam Altman asserted in a recent talk that GPT-5 will be smarter than all of us. Anthropic CEO Dario Amodei described the powerful AI systems to come as “a country of geniuses in a data center.” These are not radical predictions. They are nearly here.

But here’s the thing: these are radical predictions. Many companies tried to build the equivalent of the proposed GPT-5 and found that continuing to scale up the size of their models isn’t yielding the desired results. As described above, they’re left tuning the models they already have for specific tasks that are well-described by synthetic data sets. This can produce cool demos and products, but it’s not a route to a singular “genius” system that’s smarter than humans in some general sense.

Indeed, if you look closer at the rhetoric of the AI prophets in recent months, you’ll see a creeping awareness that, in a post-scaling law world, they no longer have a convincing story for how their predictions will manifest.

A recent Nick Bostrom video, for example, which (true to character) predicts Superintelligence might happen in less than two years (!), adds the caveat that this outcome will require key “unlocks” from the industry, which is code for we don’t know how to build systems that achieve this goal, but, hey, maybe someone will figure it out!

(The AI centrist Gary Marcus subsequently mocked Bostrom by tweeting: “for all we know, we could be just one unlock and 3-6 weeks away from levitation, interstellar travel, immortality, or room temperature superconductors, or perhaps even all four!”)

Similarly, if you look closer at AI 2027, the splashy new doomsday manifesto which argues that AI might eliminate humanity as early as 2030, you won’t find a specific account of what type of system might be capable of such feats of tyrannical brilliance. The authors instead sidestep the issue by claiming that within the next year or so, the language models we’re tuning to solve computer programming tasks will somehow come up with, on their own, code that implements breakthrough new AI technology that mere humans cannot understand.

This is an incredible claim. (What sort of synthetic data set do they imagine being able to train a language model to crack the secrets of human-level intelligence?) It’s the technological equivalent of looking at the Wright Brother’s Flyer in 1903 and thinking, “well, if they could figure this out so quickly, we should have space travel cracked by the end of the decade.”

The current energized narratives around AGI and Superintelligence seem to be fueled by a convergence of three factors: (1) the fact that scaling laws did apply for the first few generations of language models, making it easy and logical to imagine them continuing to apply up the exponential curve of capabilities in the years ahead; (2) demos of models tuned to do well on specific written tests, which we tend to intuitively associate with intelligence; and (3) tech leaders pounding furiously on the drums of sensationalism, knowing they’re rarely held to account on their predictions.

But here’s the reality: We are not currently on a trajectory to genius systems. We might figure this out in the future, but the “unlocks” required will be sufficiently numerous and slow to master that we’ll likely have plenty of clear signals and warning along the way. So, we’re not out of the woods on these issues, but at the same time, humanity is not going to be eliminated by the machines in 2030 either.

In the meantime, the breakthroughs that are happening, especially in the world of work, should be both exciting and worrisome enough on their own for now. Let’s grapple with those first.

####

For more of my thoughts on AI, check out my New Yorker archive and my podcast (in recent months, I often discuss AI in the third act of the show).

For more on my thoughts on technology and work more generally, check out my recent books on the topic: Slow Productivity, A World Without Email, and Deep Work.

The post AI and Work (Some Predictions) appeared first on Cal Newport.

Back to the (Internet) Future

On Saturday, the Washington Nationals baseball team played their first spring training game of the season. I was listening to the radio call in the background as I went about my day. I also, however, kept an eye on a community blog called Talk Nats.

The site moderators had posted an article about today’s game. As play unfolded, a group of Nationals fans gathered in the comment threads to discuss the unfolding action.

Much of the discussion focused on specific plays.

“Nasty from Ferrer,” noted a commenter, soon after one of the team’s best relief pitchers, Jose Ferrer, struck out two batters.

“Looks like we took the Ferreri [sic] out of the garage,” someone else replied.

There were also jokes, such as when, early in the game, someone deadpanned: “Anyone who K’s [strikes out] is cut.” As well as more general discussion of the season ahead.

If you followed the thread long enough, it became clear that many of the commenters know each other, while others were meeting for the first time. As the game wrapped up, someone mentions that they’re listening from a part of Canada that recently received three feet of snow. Another commentator replied by recalling a trip they took to that same area: “It was amazing.”

Ultimately, over 540 comments were left over the duration of an otherwise uneventful, early season exhibition match.

I first wrote about Talk Nats in a 2023 article for The New Yorker, titled “We Don’t Need Another Twitter.” In that piece, I was responding specifically to the launch of Meta’s Threads platform, but I had a more general point as well: perhaps it had been a mistake to try to organize the internet’s activity around a small number of massive, privately-controlled platforms, used by hundreds of millions of users all at once.

“Forcing millions of people into the same shared conversation is unnatural, requiring aggressive curation that in turn leads to the type of supercharged engagement that seems to leave everyone upset and exhausted,” I wrote. “Aggregation as a goal in this context survives…for the simple reason that it’s lucrative.”

Boutique sites like Talk Nats, by contrast, offer something closer to the original vision for the internet, which was more focused on connection and discovery; a place where a baseball fan from Canada could spend an afternoon delighting with a few dozen of his likeminded brethren about a lazy afternoon baseball game in Florida.

This is the internet as a source of joy. And it’s the opposite of the giddy paranoia or coldly-optimized numbness delivered on massive platforms like X or TikTok.

I was thinking about that New Yorker piece today as I was following the game on Talk Nats. Those ideas, it occurred to me, are even more true right now than they were when I first published them.

“I declare the global social space we are building to be naturally independent of the tyrannies you seek to impose on us,” wrote John Perry Barlow in his seminal 1996 document, A Declaration of the Independence of Cyberspace. “You have no moral right to rule us nor do you possess any methods of enforcement we have true reason to fear.”

In the thirty years that passed, we have allowed exactly this type of soul-deadening tyranny to take hold of cyberspace — an unavoidable consequence of consolidating this once distributed and quirky medium into a small number of massive platforms.

I really enjoyed my time today on Talk Nats. I didn’t come away angry or depressed, and was more uplifted than brought down. Maybe it’s time to declare independence once again.

#####

In other news…

–> For another take on this same topic, see River Page’s recent Free Press essay, “The Online Right is Building a Monster,” which does a good job of detailing the unsavory dynamics that can arise on massive internet platforms. (His critiques of both the online right and online left hit home in this one.) The solution to the woes Page documents? Stop using these services!

–> In the audio world, on Episode 341 of my podcast, released earlier this morning, I extract a lesson about the importance (and difficulty) of fighting overload in our digital world.

–> Meanwhile, as long as we’re discussing meaningful online spaces, I’ll point your attention over to The Growth Equation, where my friends Steve and Brad have posted another one of their (rightfully) famed manifestos: “How to Save Youth Sports.” [ read | subscribe ]

The post Back to the (Internet) Future appeared first on Cal Newport.

Productivity Rain Dances

A reader recently sent me a clip from Chris Williamson’s podcast. In the segment, Williamson discusses his evolving relationship with productivity:

“Look, I come from a productivity background. When I first started this show, I was chatting shit about Pomodoro timers, and Notion external brains, and Ebbinhaus forgetting curves, and all of that. Right? I’ve been through the ringer, so I’m allowed to say, and, um, you realize after a while that it ends up being this weird superstitious rain dance you’re doing, this sort of odd sort of productivity rain dance, in the desperate hope that later that day you’re going to get something done.”

I was intrigued by this term “productivity rain dance.” Some additional research revealed that Williamson had discussed the concept before. In a post from last summer, he listed the following additional examples of rain dance activities:

“Sitting at my desk when I’m not working”“Being on calls with no actual objective”“Keeping Slack notifications at zero, sitting on email trying to get the Unread number down”“Saying yes to a random dinner when someone is coming through town”

What do these varied examples, from obsessing over Ebbinhaus forgetting curves to waging war against your email inbox, have in common? They’re focused on activity in the moment instead of results over time. “The problem is that no one’s productivity goal is to maximize inputs,” Williamson explains. “It’s to maximize outputs.”

When you look around the modern office environment, and see everyone frantically answering emails as they jump on and off Zoom meetings, or watch to solo-entrepreneur lose a morning to optimizing their ChatGPT-powered personalized assistant, you’re observing rain dances. Everyone’s busy, but is no one is asking if all these gyrations are actually opening the clouds.

The solution to the rain dance phenomenon is not to abandon organizational systems or routines altogether, nor is to crudely commit to working less. It’s instead, as Williamson suggests, to turn your attention from inputs to outputs. Identify the most valuable thing you do in your job, and then figure out what actually helps you do it better. This is what you should focus on.

The answers to these questions aren’t necessarily easy. As I talk about in Slow Productivity, making more time for key efforts often requires that you first tame the less important activities that are getting in the way. You probably need a more formal workload management philosophy to avoid overload, such as using quotas or separating “active” tasks from “waiting” tasks. You’ll also need better collaboration processes that avoid the distraction of constant messaging, such as using regular office hours for complicated discussions, and some notion of time management, such as time blocking, to maintain control of your schedule.

What separates these grounded productivity efforts from productivity rain dances is that they’re not symbolic, nor are they exercises in busyness for the sake of busyness. (What I call “pseudo-productivity” in my book.) Their success is instead measured by the concrete results they produce. As a result, they’re not flashy, or high-tech, or even all that exciting to deploy. But they work.

Rain dances can be satisfying. They feel important and active in the moment, and give you all sorts of little details to tweak and adjust. But ultimately, if your goal is to reap a rich harvest, there’s no avoiding the necessity to get down among your crops, sweat on your brow, and actually work the land.

#####

In other news…

–> For an extended discussion of productivity rain dances, check out Episode #340 of my podcast.

–> If you want to see me discussing productivity with Williamson, check out my appearance on his show from last spring.

–> Over at Growth Equation, Brad Stulberg recently wrote an essay I really enjoyed: “The Case for Mastery and Mattering in a Chaotic World” [ read | subscribe ]

–> Amazon has my latest book, Slow Productivity, discounted all the way to $18.00. If you were on the fence about checking it out, this would be a good time!

The post Productivity Rain Dances appeared first on Cal Newport.

Let Brandon Cook

I recently listened to Tim Ferriss interview the prolific fantasy author Brandon Sanderson (see here for my coverage of Sanderson’s insane underground writing lair). Tim traveled to Utah to talk to Sanderson at the headquarters of his 70-person publishing and merchandising company, Dragonsteel Books.

The following exchange, from early in the conversation, caught my attention:

Ferriss: “It seems like, where we’re sitting –and we’re sitting at HQ — it seems like the design of Dragonsteel, maybe the intent behind it, is to allow you to do that [come up with stories] on some level.”

Sanderson: “Yeah, yeah, I mean everything in our company is built around, ‘let Brandon cook.’ And take away from Brandon anything he doesn’t have to think about, or doesn’t strictly need to.”

As someone who writes a lot about knowledge work in the digital age, I’m fascinated by this model of cooking, which I define as follows: a workflow designed to enable someone with a high-return skill to spend most of their time applying that skill, without distraction.

It makes sense to me that Dragonsteel goes out of its way to protect Sanderson’s ability to think and write. The roughly 300,000 words he produces per year is the raw material with which his company’s revenue is ultimately built. To significantly reduce Sanderson’s ability to produce those words might make some of his employees’ lives easier, but it would be like reducing the amount of steel shipped to an automotive assembly-line; eventually you’re going to ship many fewer cars, and your sales will plummet.

What doesn’t make sense to me is why this cooking model is so rare in knowledge work more generally. To be clear, this approach doesn’t apply to all jobs. At the moment, for example, as a full professor in Georgetown’s computer science department, I’m taking my turn as the Director of Undergraduate Studies (DUS). This is not a position built around a singular high-return skill, so it would make no sense for the department to orient around “letting Cal cook” as DUS.

But, it’s also true that there are many jobs where, like for Sanderson, letting individuals focus on a single high-return activity could really boost the bottom line. I’m thinking, for example, of programmers, researchers, engineers, and any number of creative industry positions. And yet, we almost never see something like Sanderson’s focused setup replicated.

A major culprit here is technology. Digital communication eliminates most of the friction required to command other peoples’ time and attention toward your own benefit. It costs essentially nothing to shoot off a quick message with a question, or to ask someone to jump on a call, or to pass along a task that just occurred to you.

In such an environment, in the absence of hard barriers, most people get inexorably dragged toward a degenerate equilibrium state defined by constant distraction and obligation saturation. (I’ve written two books about this effect if you want to learn more about it.) If Sanderson didn’t explicitly build his entire company around letting him cook, in other words, then he would likely find himself instead spending much of his day answering email.

What I would like to see is a world in which many organizations have, at the very least, a handful of Sanderson-type positions — employees with super high-value skills that are left alone to apply them in a focused manner. This would only impact a relatively small percentage of workers, so why would it matter? Because it would represent a notable incursion against the broader embrace of pseudo-productivity — the idea that busyness is synonymous with usefulness, and more activity is better than less. It would open our eyes to the idea that some activities are more valuable than others, and in-the-moment convenience is over-rated in the office setting. It would empower more organizations to explore more radical and interesting ways of structuring how they get things done.

I don’t need us to figure this out immediately, but it would be nice, however, if we could make some progress before my stint as DUS comes to an end. By then, I’ll for sure be ready to cook.

#####

In other news…

If you want hear more on this topic, listen to Episode 339 of my podcast, in which I discuss this topic in more detail, including more practical ideas about how to formalize and spread the cooking model.

My good writer friends Brad Stulberg and Steve Magness, over at The Growth Equation, recently published a great essay on their newsletter titled: “A Letter to My Younger Self: On Regret, Resilience, and Dealing with the Messiness of Life.” [ read online | subscribe ]

(Note: Steve also just published a great new book that I highly recommend: Win the Inside Game.)

Have you checked out my new book, Slow Productivity, yet? You should! In case it helps persuade you, it was recently revealed to be one of the top #5 most popular non-fiction books of 2024 in the Seattle library system, and the #1 most popular self-help audiobook of 2024 in the LA library system. (Wait, do I live on the wrong coast?)

The post Let Brandon Cook appeared first on Cal Newport.

The TikTok Ban Is About More Than TikTok

On Saturday night, in compliance with a law that the U.S. Supreme Court had just upheld, TikTok shut down its popular video-sharing app for American users. On Sunday, after an incoming president Trump vowed to negotiate a deal once in office, they began restoring service. It’s unclear what will happen next, as some lawmakers in the president’s own party remain firmly in favor of the divest-or-ban demand, while some democrats seemed to back-pedal.

From my perspective as a technology critic, the ultimate fate of this particular app is not the most important storyline here. What interests me more about these events is the cultural rubicon that we just crossed. To date, we’ve largely convinced ourselves that once a new technology is introduced and spread, we cannot go backward.

Social media became ubiquitous so now we’re stuck using it. Kids are zoning themselves into a stupor on TikTok, or led into rabbit holes of mental degeneration on Instagram, and we shrug our shoulders and say, “What can you do?”

The TikTok ban, even if only temporary, demonstrates we can do things. These services are not sacrosanct. Laws can be passed and our lives will still go on.

So what else should we do? I’m less concerned at this moment about national security than I am the health of our kids. If we want to pass a law that might make an even bigger difference, now is a good time to take a closer look at what Australia did last fall, when they banned social media for users under sixteen. Not long ago, that might have seemed like a non-starter in the U.S. But after our recent action against TikTok, is it really any more extreme?

It’s fortuitous timing that all of this is going down during the New Year season, when we typically think about self-improvement. Next week, for example, Scott Young and I are launching a new session of our online course, Life of Focus, which we traditionally do around this time of year. This course unfolds over three months and helps people find more depth and meaning in their work and life. Here’s what relevant to our current moment: the entire first third of the course is dedicated to digital minimalism. Scott and I realized as we were originally working on these lessons that until you repair your relationship with your devices, you won’t have the attention or energy to make a difference anywhere else.

This is why it heartens me to see our culture begin to consider stronger steps against the most powerful of digital distractions — a key instantiation of my philosophy of techno-selectionism. But you shouldn’t have to wait for the next big legislative move to begin reclaiming your autonomy from the clutches of a small number of massive online platforms. You can implement your own personal technology bans anytime you want, and there’s nothing the president, or the industry insiders who have his ear at the moment, can do to stop you.

#####

As mentioned: Life of Focus, my three-month course co-taught with Scott Young, will reopen for a new session on Monday, January 27, 2025. Find out more here.

The post The TikTok Ban Is About More Than TikTok appeared first on Cal Newport.

Lessons from YouTube’s Extreme Makers

In 2006, a high school student from Ontario named James Hobson started posting to a new platform called YouTube. His early videos were meant for his friends, and focused on hobbies (like parkour) and silliness (like one clip in which he drinks a cup of raw eggs).

Hobson’s relationship with YouTube evolved in 2013. Now a trained engineer, he put his skills to work in crafting a pair of metal claws based on the Marvel character, Wolverine. The video was a hit. He then built a working version of the exoskeleton used by Matt Damon’s character in the movie Elysium. This was an even bigger hit. This idea of creating real life versions of props from comics and movies proved popular. Hobson quit his job to create these videos full-time, calling himself, “The Hacksmith.”

Around the same time that Hobson got started on YouTube, a young British plumber named Colin Furze also began experimenting with the platform. Like Hobson, he began by posting videos of his hobbies (like BMX tricks) and silliness (like a stunt in which tried to serve food to moving cars).

Furze’s relationship with YouTube evolved when he began posting record breaking attempts. The first in this informal series was his effort to create the world’s largest bonfire. (“I collected pallets for over a year.”) He drew attention from British media when he supercharged a mobility scooter to drive more than seventy miles per hour. This led to a brief stint as a co-host of a maker show called “Gadget Geeks” that aired on the then fledgling Sky TV. After that traditional media experience, he scored a hit on YouTube by attaching a jet engine to the back of a bicycle. He decided to fully commit to making a living on his own videos.

I wrote about Hobson and Furze in my most recent essay for The New Yorker, which was titled, “A Lesson in Creativity and Capitalism from Two Zany YouTubers.” What drew my attention to these characters, and provided the main focus for my article, is what happened after they decided to make posting videos their full-time jobs.

Hobson adopted a standard strategy from the media industry: he tried to grow as fast as possible. He moved from his garage to a leased warehouse, and then, when that lease ran out, he took on a multi-million dollar mortgage to buy an even larger warehouse. He soon had thirty employees and around a quarter million dollars a month in overhead.

Furze, by contrast, stayed small. He continued to film his videos in his home workshop and a nearby old barn. He worked almost entirely on his own, with the exception of sometimes having his wife help hold a camera, or his friend Rick come lend a hand when some extra strength was needed. Furze’s overhead was reduced to more or less the cost of materials. Everything else he earns he keeps.

Hobson and Furze’s opposite strategies provide a neat natural experiment in the economics of this quirky corner of YouTube. What were the results? In 2024, Hobson’s channel published twenty-five beautifully produced videos that attracted more than twenty-seven million total views. In the same period, Furze launched five solo-produced videos on his main channel that attracted eighteen million views. He also, however, maintained a second channel with behind-the-scenes footage that pushes his total views for the year to forty-three million, nearly double Hobson’s results.

As I write:

“Furze’s solo success is a quirky challenge to the traditional narrative that survival requires continually growing, and that a small number of well-financed winners eventually eat most of the economic pie. He demonstrates that in certain corners of the creative economy an individual with minimal overhead can work on select attention-catching projects and earn a generous upper-middle-class income. Beyond this relatively modest scale of activity, however, the returns on additional investment rapidly diminish. As Hobson’s experience suggests, there’s no obvious path for a D.I.Y. video creator to turn his channel into a multimillion-dollar empire, even if he wants to. Furze seems to be maxing out the financial potential of his medium by staying small.”

In my article, I go on to the explore the specific reasons why small works so well in this medium (hint: it has to do with maintaining an authentic personal connection with your audience). But what I want to emphasize here is my broader conclusion. I think these particular corners of YouTube, along with some related creator-focused Internet-based technologies, including emails newsletter and podcasts, are helping to carve out space for a relatively broad “creative middle class.”

As social media continues to falter and stumble in its role as a unifying cultural force, its model of people volunteering their creative labor in return for uncompensated attention is beginning to lose its appeal. Colin Furze is one among many who are revealing an alternative engagement with the online world; one in which it’s possible for someone with sufficient talent to make a good living with minimal investment and maximal flexibility.

As I conclude in my piece, it’s still really hard to succeed in this new creative economy. But at least there’s space now to do so. As I write:

“In our era of consolidation and polarization, many online spaces can seem dreary, toxic, addicting, or some combination of the three. As my colleague Kyle Chayka wrote in 2023, most of the Web just ‘isn’t fun anymore.’ In Furze, however, I sensed some of the optimism of the early Internet.”

Sounds good to me.

#####

In Other News…

For nearly two decades, my friend Adam Gilbert (featured here in a 2007 Study Hacks post) has run My Body Tutor, an immensely successful health and fitness app that is based on the simple but powerful idea of using online coaches to hold people accountable.

His team just launched a new platform called DoneDaily that brings this same coach-driven accountability to professional productivity. I’m mentioning it here because DoneDaily deploys a lot of ideas I talk about here and in my books — including, notably, multi-scale planning — but now combined with a dedicated coach who you check in with daily to make sure your plan makes sense and that you’re taking action.

Anyway, I thought this was one of those ideas that makes so much sense that it’s surprising it didn’t exist before. Indeed, it’s the type of thing I might have built on my own if I didn’t already have a bunch of jobs. So I’m glad Adam got there first and was happy, at his request, to help share it. Check it out!

(Note: I have an affiliate relationship with this site.)

The post Lessons from YouTube’s Extreme Makers appeared first on Cal Newport.

The Tao of Cal

Between this newsletter, my podcast, my books, and my New Yorker journalism, I offer a lot of advice and propose a lot of ideas about how the modern digital environment impacts our lives, both professionally and personally, and how we should respond.

This techno-pontification covers everything from the nitty gritty details of producing good work in an office saturated with emails and Zoom, to heady decisions about shaping a meaningful life amid the nihilistic abstraction of an increasingly networked existence.

With the end of year rapidly approaching, and people finding themselves with some spare thinking time as work winds down for the holidays, I thought it might be fun to try to summarize essentially every major idea I discuss in one short primer.

So that’s what I’m attempting below! I’m sure I’m missing some key points, but this should nevertheless provide a useful road map to my esoteric mental wanderings.

Knowledge Work

Treat cognitive context shifts as “productivity poison.” The more you switch your attention from one target (say, a report you’re writing) to another (say, an inbox check), the more exhausted and dumber you become.The biggest source of context shifts is digital communication. Move as much collaboration as possible out of chains of ad hoc, back and forth messaging and into something more structured.The second biggest source of context shifts is caused by working on too many tasks at the same time. Do fewer things at once. You’ll finish each task faster (and better) and therefore accomplish more over time.Focus is like a super power in most knowledge work jobs. Train this ability. Protect deep work on your calendar. Support these sessions through special rituals and spaces.You need specific systems to track all of your commitments. You need specific system to manage your time and attention. The pace and volume of modern knowledge work is too intense for you to casually handle it all in your head.Remote work requires more structure surrounding workload management and communication than regular office work. It’s not enough to simply give remote workers a Zoom account and a Slack handle and hope their efforts unfold as before.

Sources: Deep Work, A World Without Email, Slow Productivity, “Why Remote Work is So Hard–And How it Can Be Fixed”, “Why Do We Work Too Much?”, “Was Email a Mistake?”, “How to Have a More Productive Year”

Personal Technology UseYour phone should be used as a tool, not a constant companion. To accomplish this: (1) keep your phone plugged into the same spot when at home (instead of having it with you); and (2) remove all apps from your phone where someone makes more money the more you use it.Most people don’t need to use social media. If you really need to use it — e.g., for professional purposes — use it on a web browser on your laptop, and spend at most an hour a week logged in, as that’s enough for 99% of legitimate uses. There are better ways to be entertained, find news, and connect with people.Digital communication can be great, but be wary of communicating with people you’ve never actually met in person before. (That is, texting a friend is good. Arguing with a random Twitter user about presidential politics is not.)Fixing your relationship with digital tools requires that you fix your analog life first. It’s not enough to stop using problematic apps and devices, you must also aggressively pursue alternative activities to fill the voids this digital abstention will create: read books, join communities, develop hard hobbies, get in shape, hatch plans to transform your career for the better. Without deeper purpose, the shallow siren song of your phone will become impossible to ignore.Kids under the age of 16 shouldn’t have unrestricted access to the internet. Their brains aren’t ready for it.

Sources: Digital Minimalism, “Quit Social Media”, “Steve Jobs Never Wanted Us to Use Our iPhones Like This”, “Cal Newport on Kids and Smartphones”

The Deep LifeIn building a meaningful and fulfilling life, it’s usually better to work backwards from a broad vision of your ideal lifestyle than it is to work forward toward a singular grand goal (e.g., a “dream job” or radical location change) that you hope will make everything better.The best way to improve your professional life is to get good at something the market unambiguously values, and then use this “career capital” as leverage to shape your work in ways that resonate. No one owes you a great a job. You have to get great first before you demand it.Succeeding with big changes in your life requires that you first get your act together. Get comfortable with discipline (doing things that are hard in the moment but important in the long term), get organized, and reclaim your brain from constant digital distraction. Only then should initiate your ambitious plans.

Sources: So Good They Can’t Ignore You, “The Most Important Piece of Career Advice You Probably Never Heard”, “The Deep Life: Some Notes”, “Deep Life Stack 2.0”

The Internet and Future TechnologyWhen it comes to the internet, small is usually better than big. Niche online communities are more meaningful and less harmful (in terms of both content and addictive properties) than massive social platforms. Independent content formats, like podcasts and newsletters, are much better for creatives (in terms of stability, income, and autonomy) than attempting to become an influencer on a major platform. And so on.The age of massive social network monopolies is already coming to an end. We just don’t realize it yet.Generative AI won’t really change our daily lives in a massive way until it leaves the chatbot format and becomes more integrated into specific tools.The biggest technology story everyone is ignoring is the end of screens. Within the next decade, AR glasses will replace essentially every screen currently in our lives — phones, laptops, tablets, computer monitors, and televisions. The ramifications on the worldwide technology sector will be absolutely massive. It will also be the end of a fully differentiated analog reality as we know it.

Sources: “The Rise of the Internet’s Creative Middle Class”, “TikTok and the Fall of the Social Media Giants”, “Can an AI Make Plans”, “The End of Screens?”

#####

Speaking of books, my latest, Slow Productivity: The Lost Art of Accomplishment without Burnout , was named a best book of 2024 by The Economist , NPR , and Amazon . It’s also currently heavily discounted for Cyber Monday. Consider it as a great gift for yourself or someone else you know who could benefit from slowing down! (The photo used for this article was taken by Greg Kahn for a recent profile of me published in El Pais.)

The post The Tao of Cal appeared first on Cal Newport.

After You Vote: Unplug

I’m writing this post about eighteen hours before the first polls open on Election Day, and it feels tense out there. The New York Times, for example, just posted an article headlined: “How Americans Feel About the Election: Anxious and Scared.”

Based on extensive interviews conducted over this past weekend, the Times concludes:

“Americans across the political spectrum reported heading to the polls in battleground states with a sense that their nation was coming undone. While some expressed relief that the long election season was finally nearing an end, it was hard to escape the undercurrent of uneasiness about Election Day.”

These results probably come as no surprise.

The question then becomes what to do with this anxiety. The first step, of course, is to vote — and not just vote, but to approach your decision honestly and dispassionately. By the time you read this, you’ve likely already completed this step.

But then what?

Here I have a suggestion that I think could be healing for all points of the political spectrum: use the stress of this election to be the final push needed to step away from the exhausting digital chatter that’s been dominating your brain. Take a break from social media. Stop listening to news podcasts. Unsubscribe, at least for a while, from those political newsletters clogging your inbox with their hot takes and tired in-fighting.

I suggest you switch to a slower pace of media consumption. Don’t laugh at this suggestion, because I’m actually serious: consider picking up the occasional old-fashioned printed newspaper (free from algorithmic optimization and click-bait curation) at your local coffee shop or library to check in, all at once, on anything major going on in the world. I think I might setup a Sunday-only paper subscription as my main source of news this winter.

Equally important is how you redirect your newly liberated attention. Consider aiming it toward real community, with real people who actually live near you, to retrain your brain to stop thinking of the world as hopelessly fractured into vicious tribes. (If right now you’re scouring this post to seek evidence as to whether I’m friend or foe, then you’re already severely suffering from this malady. )

Consider reading books again. There’s a pleasure in the conquest of deep ideas that’s been lost as we thrashed in a digital sea of churning distraction. Spend more time in nature to discover that despite the apocalyptic tenor of the online world, its analog counterpart persists, and is beautiful.

The Republic will still stand without our constant digital vigilance. But it’s unclear if our mental health can survive the status quo.

#####

Two announcements to share…

For the past twelve years, my longtime friend Joshua Fields Millburn, of The Minimalists fame, has been teaching a fantastic online course called How to Write Better. It’s open this week for a new session, so if you’re interested in improving your writing ability (which you should be), please check it out! If you’re still on the fence about whether or not to read my new book, Slow Productivity , check out this insightful new review from Real Clear Books that was published last week.

The post After You Vote: Unplug appeared first on Cal Newport.

The Perfect Cheating Machine?

Many predictions and concerns tumbled into the slipstream trailing ChatGPT’s dazzling, turbulent entrance onto the technology scene in late 2022. Few of these initial warnings felt more immediate than those of imminent disruptions to higher education.

“Could the chatbot, which provides coherent, quirky, and conversational responses to simple language inquires, inspire more students to cheat?”, asked an NBC News article, published only a week after ChatGPT’s initial launch. Several months later, a professor in the Texas A&M system took this warning to heart and failed his entire class after convincing himself that every one of his students had used AI to write their final assignments. (It turns out that his method of detection—asking ChatGPT itself whether it produced the submissions—was unreliable. He later changed the grades.)

“AI seems almost built for cheating,” explains Ethan Mollick, in his recent bestseller, Co-Intelligence. He predicted, in particular, that paper writing as a pedagogical tool might be on the way out, forcing institutions to adapt to other methods to teach composition: “In-school assignments on non-internet-enabled computers, combined with written exams, will ensure students learn basic writing skills.”

It’s hard to believe that it’s been almost two years since we first started hearing these concerns about ChatGPT providing students the perfect cheating machine. As a professor and writer myself, these issues interest me, especially when it comes to academic compositions. So in my most recent article for The New Yorker, published earlier this month, and titled “What Kind of Writer is ChatGPT?,” I set out to understand how these tools are currently being put to work by students tackling writing assignments.

My approach was to move beyond speculation and watch actual students use AI on actual assignments, with a particular focus on a graduate student I called Chris, who was using ChatGPT to write a significant anthropology paper.

As I explain in the article, what I observed Chris doing was more complicated than you might have guessed:

“He was not outsourcing his exam to ChatGPT; he rarely made use of the new text or revisions that the chatbot provided. He also didn’t seem to be streamlining or speeding up his writing process. If I had been Chris’s professor, I would have wanted him to disclose his use of the tool, but I don’t think I would have considered it cheating. So what was it?”

I recommend that you read the full article to learn the full answer. But to preview what I discovered: students aren’t simply outsourcing their writing to tools like ChatGPT, but they’re also not using them in clearly harmless ways either. The reality is something different and new; less a method to speed up the task of writing and more an approach to reducing its cognitive burden.

The bigger point to be made here, however, is about how we think about this new age of artificial intelligence in which we’ve been enveloped since late 2022. These tools are undeniably powerful. Accordingly, they will undeniably end up changing some things about our lives in major ways.

But predicting these changes has proven exceedingly difficult. If you’re interested in these trends, spend less time listening to people explaining how the next version of some model is going to change everything all at once, and instead directly observe what people are doing with the versions of the technology they have access to right now. The stories are less flashy, but as you look deeper you’ll find interesting things going on.

#####

In the latest episode of my podcast, Deep Questions, released earlier today, I take an unconventional look at the idea of discipline and how to improve it (listen | watch)

Have you read my new book, Slow Productivity: The Lost Art of Accomplishment without Burnout? If not, here’s yet another reason to consider doing so: Goodreads just listed Slow Productivity as #1 on their recently released list of the “Most Popular Self Help Books of 2024.”

The post The Perfect Cheating Machine? appeared first on Cal Newport.

When Time Management Was Easy

In 1973, an author named Alan Lakein published a book titled How to Get Control of Your Time and Your Life. It wasn’t the first book about professional time management — my library contains a first edition of James McCay’s 1959 classic, The Management of Time — but it’s arguably the first book to talk about the topic in a recognizably modern way, with a focus on personalized tools like daily to-do lists. It went on to reportedly sell more than three million copies, and was even shouted out by Bill Clinton, who cites its influence on his early career in his autobiography.

Revisiting Lakein’s advice today provides a glimpse into office life fifty years ago. And the encounter is shocking.

One of Lakein’s more famous suggestions is to write down everything you need to do on a single task list. He then says to label each task with one of three priorities: “A” for things that are important and urgent, such as those with impending deadlines; “B” is for tasks that are important but not urgent, and can therefore be postponed if necessary; “C” is for things that are small, easy, and don’t require attention at the moment.

You start by completing the A tasks, crossing them off your list as you go. Then you move on to the B category. If you finish the B tasks, you can tackle some of the C. Lakein notes that these task priorities might evolve. An important obligation with a distant deadline, for example, might start at B, but then, as the deadline approaches, upgrade to A. Lakein’s intention is to help you make sure that you make progress on the things that most require your attention.

Part of what’s shocking about this system is its finitude. In 2024, can you imagine fitting everything you need to do on a single list? Your email inbox alone could likely contribute several hundred items at any given moment. Also notable is Lakein’s assumption of task stability; that your list would more or less stay the same as you carefully worked your way through it during a full workday. Modern work is instead defined by constant new demands — chats, questions, meeting invitations, requests to “jump on a call” — that require timely answers.

Here’s the question that began to fester as I revisited these older ideas: is what we’re doing today any better?

The fact that our modern workflows would swamp Lakein’s quaint system of simple lists and priorities is perhaps more an inditement of us than him. To have more work, arriving with much more urgency, than we can possibly get our arms around is not a good recipe for getting useful effort out of human brains. It is, however, a good recipe for burnout.

A point I often make on my podcast, as well as in my new book, Slow Productivity, is that in my own work on these topics, I describe more complicated time management strategies with reluctance. My bigger wish is to help reform office work to the point that they’re no longer needed, and something like Lakein’s basic ABC system is more than enough.

We’re not there yet, but in the meantime, it helps to realize where we are now isn’t working.

#####

If you want to learn more about what’s (regrettably) required to manage your time in our current moment, check out the latest episode of my podcast, Deep Questions , where I tackle three advanced time techniques ( listen | watch ).If you haven’t yet read my latest book, Slow Productivity, you should! Some more encouragement: (1) it was recently named an official selection of The Next Big Idea Club [meaning it was chosen by a panel consisting of Malcolm Gladwell, Adam Grant, Dan Pink, and Susan Cain as one of the two best idea books of the season]; and (2) it was selected for the shortlist for SABEW’s Best Business Book of 2024 award.

The post When Time Management Was Easy appeared first on Cal Newport.