Ask Paul: November 22 (Premium)

Happy Friday! We’re back in the US and exhausted, but here’s a particularly lengthy installment of Ask Paul to kick off the weekend a bit early.

Pretty no bueno

j5 asks:

Hi Paul! Any updates on when your Paulism “pretty bueno” is going to be on your coffee mugs? That’s what I’m waiting for to buy one. Thanks!

Sorry, this hasn’t really been top of mind lately. I did update the swag store with several more items at some point a few weeks or maybe a month back, but that wasn’t among them. I’ll do that over the weekend.

Speaking of which.

This reminds me of something that happens each year on Windows Weekly, mostly because it just came up Wednesday, when Leo asks me if I have any favorite moments from the year and I never do. To me, this is a blur of 2-3 hours every Wednesday and I don’t actually hang on whatever stupid phrases come out of my mouth, though some are memorialized as episode titles. In this case, I know there are likely some other phrases that might be fun to have on t-shirts or mugs or whatever. But I don’t remember a single one of them. If anyone has thoughts on this, let me know.

(OK, I can think of one, but maybe it’s only funny/relevant to me: “But you said…”)

Chicken Little

wright_is asks:

I hear a lot of people calling out the sanctions being suggested by the FTC against Google (and other bodies against Google, Apple, Meta, Microsoft, Amazon etc.), complaining that it might cause the company to “fail” or to have considerably reduced income etc. Surely that is the point? The sanctions are there to punish the companies and to ensure, going forward, that they comply with the law.

This is correct. Google was found to have an illegal monopoly, meaning it behaved illegally to create or maintain that monopoly, and the remedy must be designed to not just curb that behavior, but to prevent future abuses and to deny Google of its illegal revenues and other benefits. We can absolutely debate the merits of whatever proposed remedies the DOJ/states come up. But it’s fascinating to me that this could have–maybe should have–been much worse. For example, why not just cut off the head of the snake and force Google to divest itself of what used to be DoubleClick, i.e. its ad business? That’s the simplest way to achieve the goals. But that’s also a lot more radical than the proposals the DOJ suggested. Why isn’t the DOJ being praised for a measured set of proposals?

For me, it is simple, if a business can’t survive on its merits, it isn’t a viable business. If the business only survived due to illegal practices and can’t survive without them, surely that is the natural order of things, that those businesses cease to be?

Also correct. This is what we can now call the Mozilla Paradox, I guess. Because in addition to Google, which I think most would agree is this overly powerful and abusive Big Tech monopoly, we have collateral damage in the form of Apple, Mozilla, and Samsung, each of which will see its revenues go down in a material way when Google is forced to end the payola. No one cares about Apple and Samsung–well, beyond investors, I guess–but Mozilla is a tough one. Even those who don’t use Firefox probably have a soft spot for this company, though it’s also made some horrible business decisions. But it is almost certainly being kept afloat artificially by the payments Google makes to keep using Google Search as the default And it will almost certainly disappear, or become something much smaller and different, when that ends.

Mozilla knows this. It’s been downsizing all year. But to your point, yes, if this thing can’t survive organically, it should not survive. It’s a business, not a small child. Whatever the business version is of Darwin’s Law, it applies. And you know what? Even the people who work for Mozilla understand that.

We’ve seen it in the past, the East India Company, Standard Oil and many others over the centuries, they were all “too big to fail”, yet they broke the law, were cut down to size and it turned out that their business models weren’t viable, without breaking the law and they ceased to be.

As you alluded to, Google won’t fail even if the judge accepts all the DOJ/states’ suggested remedies. In the most recent quarter, it earned $22.4 billion in revenues from non-advertising-based businesses (Google Cloud, platforms, devices, subscriptions, etc.), so that’s a run rate of almost $100 billion a year. But most crucially, no one is suggesting–for some reason–that Google be forced to divest its ad businesses. And that means they will survive past whatever happens. Will they be diminished? Sure. It will be a more competitive market. But by how much? Is 50 percent aggressive? Maybe. But let’s use that. 50 percent of the $65.9 billion in revenues Google made in the most recent quarter is about $33 billion. So even in a worst-case scenario, Google emerges on the other side of this with over $55 billion in revenues. In one quarter. That’s over $200 billion in a year. (And doesn’t factor in how much it will make selling off businesses and other assets.)

If you look at the list of the biggest companies on earth by revenue, that moves Google from #18 down to #30/31. Ahead of Samsung, Foxconn, Ford, Bank of America, Mercedes-Benz, and Citigroup. I mean. Boo-hoo. Poor Google.

Society is still here, society will survive if one of these illegal monstrosities cannot survive, when they are forced to follow the law. I’m sure there wouldn’t be the same outcry, if it was forcing a drug cartel to only sell its drugs legally in shops, with prescriptions and it wasn’t allowed to kill its competitor (literally) or law enforcement officers. Yet, when a business on the Stock Exchange does something illegal, especially when the company has developed a “cult of personality” for its leadership and products, many people say they should have a pass and be able to do what they want… It just doesn’t make sense to me.

Also correct. (This is something to keep in mind when it comes to AI and job losses: All new technology–steam trains, planes, computers, the Internet, whatever–is accompanied by job losses. But we look back years, decades, or generations later and … don’t think about it.) No one today cares that Standard Oil or AT&T were broken up, that’s part of history, but it was a big deal at the time. Things change. I feel like that’s especially true of the technology sector. You’d think we’d be used to that, but it’s odd that such short-lived things become traditions and that we seem unwilling to change.

There is a lot of overly simplistic thinking out there, which I understand. But these are always the loudest voices, too, and the fringe opinions most often repeated. This notion that Google, Apple, or whatever company should be able to do what they want. Or if a company doesn’t like the laws of some jurisdiction, they should just stop selling products there. Or that a company will be destroyed because we finally, belatedly curb its illegal behaviors. It’s not just uneducated, it’s sad. Life is full of nuance, but understanding right and wrong, that people matter more than businesses, that companies have been allowed for too long to become too abusive, is just common sense. It’s not political, the ultimate in failed arguments. It’s just common sense. We all know right from wrong. And we all know when we’re looking the other way for ulterior reasons. Pretending otherwise is a lie.

We can hold complex thoughts in our heads. We can understand that Google Search, say, is the best search engine, but also understand that Google got there illegally and is killing innovation elsewhere to keep it there. And that the former does not outweigh, or matter more than, the latter. That breaking the law obviates any arguments about whatever benefits.

Is there a single example of a company that was broken up that we look back on later and say, wow, that was a mistake? I can’t think of one. It’s far more likely that we’ll look back on this Google issue and only wonder one thing. What took so f#$king long?

Glucose

beewacker asks:

In the not-too-distant past you were using a glucose monitor, which your doctor prescribed. Are you still using it and do your eating habits change when you’re in Mexico vs at home? Is it something you’d recommend in lieu of the current Ozempic craze?

Yes, definitely. The only issue with glucose monitors is getting them: If you live in the US, you have some of the worst health care and health insurance in the world, and you have to pay out of pocket for this stuff because it’s preventative. If you have diabetes, it’s free (or, low-cost, really). If you are trying to prevent diabetes, as I am/was, you can’t get it. Great system.

I used a glucose monitor for a couple of months in 2023 specifically to monitor what happened when I ate different foods. I went into this knowing that some foods–bagels, most notoriously, pasta, and other carbs, starches, and sugars–can spike blood glucose dangerously, and that continually subjecting your body that effect causes lasting damage. But I also knew that the future of health care was personalized because we’re all different, and we all process foods (and components of foods) differently. We know this because two different people who eat exactly the same thing will be different weight and have different health outcomes (including twins, interestingly). And so I wanted to test some theories and see whether some specific foods that are on the bad side of the carb/glucose spike side of the chart might somehow not be problematic for me.

What I found was that all the classic carbs, sugars, and starches are problematic. The one exception, which isn’t really an exception but is instead a known scientific side effect, is twice-baked potatoes. If you cook a potato twice, letting it cool down to room temperature first before reheating it, your glucose doesn’t spike as bad as it would otherwise. (It still spikes a bit, but not as bad.) This isn’t true of bread or pasta, just potatoes, and it has something to do with those starches and how they change chemically when heated. I don’t understand it, I just know it to be true.

Anyway.

I don’t use a glucose monitor regularly because it’s expensive for me as a non-diabetic American. I always figured I’d return to it for a month or whatever every year or so, and I would still like to do that. I did lose a lot of weight by adjusting my diet in 2023 and into early 2024. Too much, at one point: When we came back from Mexico in early March, I was actually kind of gaunt, and when my sister saw me for the first time, she gasped out loud. So I kind of over-corrected on the food and reintroduced some carbs and gained some weight back. But more than I’d like, maybe 15 pounds or so. So I’ll be adjusting again after the holidays. But I’m not pre-diabetic anymore, I ran lab tests before this previous trip to Mexico, and it’s all good, and I feel like just keeping up on this is fine.

Regarding Ozempic and pills like that, I love that it exists and that it helps people. But the problem with these solutions is that you have to keep taking the pills. If you stop, you’ll gain the weight back unless you also make changes to how you eat. So it’s a sort of accelerant, I guess, and given how many people are obese these days, maybe necessary. But it has to be done alongside actual change. If you want to lose weight for good, with some exceptions–again, we’re all different–you gotta do the work too. And knowing what impacts you the most, glucose spike-wise, is useful information. A glucose monitor works well for that.

I think this is moving towards being something you can get over the counter. That will be a good day. As with everything else–see the above conversation about antitrust–education and knowledge is key. To understand our health, we need the full picture. And that can be a bit different for everyone.

Just asking quAIstions

Markld asks:

I truly am pondering about the future of AI and its impact on all of us. I appreciate your input so far as you have been [at least to me] asking or commenting on what is important. You addressed AI on the other day with “Agentic (Premium)” as an example. You/Thurrott.com have not left this subject alone which is much appreciated. I think you recognize AI is a powerful tool with potential benefits and risks, and you know thoughtful use is key. You are not hysterical, nor hype, but you keep a healthy skepticism mixed in with being pragmatic in A.I. matters.

It’s difficult to understand what the long-term impact of AI will be since we’re living through this fast-moving change right now. And we don’t know how it’s going to turn out. I trend towards the center on all things, but with AI, it’s pretty obvious that the two extremes–it’s nonsense, and, it’s going to destroy humanity–are both ridiculously naive stances. It’s going to land somewhere in the middle. I certainly went through the seven stages of grief, or whatever, with AI, just trying to understand what was happening, as it was happening. And I can’t claim to fully understand where it’s going to end up. I mean, no one can.

I have so many questions and I am not sure I want them to be answered, as much as I want to come up with questions we need to address now and/or on an ongoing basis. [I guess I am looking for just questions for all of us to be asking.

I think about the ethical and societal impacts of AI, is it going to make us less intellectual or intelligent? [I think society’s intellect is important] I have a big concern for AI and its unintended consequences, too. So that appears to be my theme for these questions below:

• “How can AI be developed and used to benefit everyone while avoiding ethical risks and unintended consequences?”

I can only speculate here. But it’s interesting that ethics is so top of mind with AI. I wonder if it’s because of the science fiction that predated that, from Asimov’s Robot stories to HAL 9000 in 2001 to the Terminator movies. That we accepted that this is what AI is, and Microsoft–or whomever–is up on stage talking about a feature that will help you summarize an email or whatever, and all we can imagine is a killer robot hunting down human victims.

What are the ethical concerns here, really? That AI will lead to intellectual property theft? Technology has always made that easier. The Internet lets us easily push stolen content to the masses. Photoshop helps us fake photos, in essence telling a lie. We can all self-publish, meaning that people with uneducated opinions have a forum and that enough of these things, in total, leads to a lot of contradictory information. I’m not sure what it is about AI that creates such angst–look at the silly, alarmist reactions to Recall for an obvious example–while ignoring the benefits.

Companies like Microsoft, Google, and others will push the ethical bit for marketing purposes. But the bigger issue to me is accuracy, ensuring that AI is trained on vetted, known-good data sources. And that it not become some weird technical form of cancer where AI is trained on AI that was itself created using lousy data. Which has, of course, already happened.

• “How can AI enhance our intellect and creativity without making us too dependent on it?”

There’s an Asimov story in the Robot series that features a planet with just several people living on it, all separated by hundreds or thousands of miles, each alone, living in some huge mansion, and communicating with the others via giant screens. (It’s like the vision for Skype.) There’s also a more simplistic, modern version of this vision of the future in the Pixar movie WALL-E, where humans are all fat and lazy, and need to move around on little floating discs because they can no longer care for themselves. But the flipside to these caricatures is that AI, or technology in general, could also be seen as the thing that helps us lead better lives, more fulfilling existences. We all work because we have to, but what if we didn’t have to? Or what if we didn’t have to settle on a monotonous job because we needed it, but could instead choose to do more interesting work?

(In the world of WALL-E, it occurs that a society that advanced would be healthy, not fat and lazy, since they would have solved those problems long ago.)

I think about this like retirement. People envision sitting on a beach with umbrella drinks, but the reality is that such a thing is boring over time, and we all need some purpose in life, some reason to get up every day. I’m never really retiring, but I could see working less and focusing on those things that matter most to me. And at a very high level, the way I might get from here to there is by being financially independent, which would give me the ability to say no to things I now do because I have to. Does this mean I’m more dependent on money, I guess, or technology? Or does it mean that my life is now less stressful and more meaningful?

We all have unhealthy relationships, with people and technology. But that’s something we can work on and improve. Finding balance is always the goal, with AI or whatever else.

• “How will AI change our jobs, and how can we prepare for working alongside it?”

This is job-specific.

I work in an industry that’s based on writing, and I am surrounded by horrible writing. To date, that horrible writing has come from people who, like me grew up in an English-speaking place and still don’t know their own language well. It’s evolved to include less experienced writers in other countries for whom English is a second language, and some of them take terrible writing to a new low (like Neowin, it’s unbelievable). And it is evolving right now to include AI. We can see progressions–digressions, I guess–in almost any industry. Apple moving all device manufacturing to China and now looking to go even cheaper in places like Vietnam and India. Next will be Africa, I guess. And then what? Mars? How far can they go?

Anyway, I came up in a world that was all paper-based publishing. Now it’s all digital. Things change. Things will keep changing. It’s on us to stay educated and informed, to read and improve, and to roll with the changes. If AI steals your job, you’re either not doing good enough work or that job shouldn’t exist anymore. We will look back and wonder why people like me were ever able to make money doing what I do. That’s just being self-aware.

• “What skills should we focus on that AI can’t easily replace, like creativity and empathy?”

So, I think AI can replace, or at least help with, these things. The creativity bit is obvious enough, an AI-created painting recently won an award. But even empathy is possible. Pre-AI Alexa devices were being used by otherwise lonely older people who had perhaps lost a spouse or were otherwise alone much of the time, by giving them something to interact with. Something that doesn’t need to be fed or cared for like a pet. Maybe this isn’t a replacement for other forms of interaction. But it is one solution for that.

• “How can AI be designed to complement our abilities and encourage working together, rather than making us reliant on it?”

This is related to an earlier question, but this is like using photographic equipment or Photoshop to preserve or enhance images that are memories and would otherwise be lost or would take too long to create with paint or whatever. Just as a writer should use whatever spelling and grammar tools that are available to them, he should these days be looking at AI to see how or whether that can improve their writing. Being obstinate about this–as I can be, actually–is counterproductive.

• “What specific steps can we take to make sure AI helps us work better and not just differently?”

This isn’t entirely on us. When I wrote I Will Not Pay for AI (Premium), my point wasn’t that I wouldn’t use AI, it was that marketing AI as AI was silly, and that for AI to be truly useful, it would just be added to those things I do pay for when it made sense to do so. Since then, Microsoft has started shifting towards a model where customers pay for Microsoft 365, which makes sense to me, and away from customers also paying for AI features in a separate Copilot product. AI features are not the thing they are a part of that thing. This makes sense to me.

• “How can AI improve areas like healthcare and the environment while keeping people at the center of its advancements?”

You may have seen the story recently that AI diagnoses are already more accurate than those provided by human doctors. This is another example of AI making something better, and freeing those people (doctors) to spend their time on more rewarding work, which ideally includes spending more time with their patients. Maybe this will help lower health costs by removing or at least reducing malpractice suits since there won’t be anyone to sue and there won’t be any malicious intent. Anything to make health care in this country better would be an improvement.

• “What are the most exciting areas of AI research that could change our lives in the next decade?”

Curing cancer and other terminal diseases. Deciphering historical documents and uncovering new information about the past. Overcoming scientific obstacles in general to make advances that otherwise might have never occurred or would have taken much longer. There are so many ways in which AI could be used to benefit mankind, and broadly.

Or, maybe there will be a robot uprising. 🙂

• “How can AI create personalized learning experiences that fit each person’s strengths and needs?”

Also personalized healthcare, per the earlier question. This is the type of thing that will happen when AI is grounded on an individual. I’m still thinking in terms of fairly immature and unsophisticated AI understanding–an AI that is grounded on all the writing I’ve done, as stored in OneDrive or whatever–but this will require a more holistic and broad grounding. We live in a day in which people are rejecting AI that’s basically a glorified spell-checker, so there are hurdles. But it’s inevitable.

To put it back in the healthcare example, we’re already seeing doctors referencing Google Search to aid with diagnoses, and that will shift to AI, hopefully AI that is grounded in accurate health data that would otherwise be hard for them to read and parse. When my son was one, he almost died from bacterial meningitis, and he lucked out because the doctor suspected it despite that not being common at the time. Another doctor would have sent us home with some medicine or whatever. In the future, those kinds of diagnoses will be more common and reliable because of this AI grounding and usage, hopefully. Certainly, that’s the goal. And you can apply that kind of thing to almost any aspect of life. Think about something as simple/basic as a book recommendation, which today amounts to, you liked this, maybe you’ll like that too. These things will all become more sophisticated because the AI will no more about you, not just that you bought one book.

The biggest hurdle to all this is understanding what AI can and cannot do. This is shifting. But again, I think the most alarmist views are way off, as are the most sanguine views. This is a big deal, but so was the Internet and the GUI for PCs. Or air travel. We can fly into space and back, but has it changed your life personally? Has it changed the day to day?

So many questions. But all we can do is speculate, aided in part by the experiences of the past. I don’t see an AI apocalypse for mankind. I see life getting better overall, but also the short-term displacement that comes with any change.

Which Apple services

helix2301 asks:

I know you and Brad talked about Apple Services business on FRD other day. My question is where do you think all that’s coming from besides Google.

Apple doesn’t break down its Services business that transparently, but I assume the biggest revenue contributors are the App Store, iCloud, and then Apple Music, in that order. The App Store is definitely the biggest component of that. And it’s much bigger than whatever they get from Google.

Netflix does more video traffic in a day then Apple does in a month on all Apple TV + content according to new articles Leo talked about on Macbreak. Apple will start licensing their TV content to other services to make it profitable.

Yes. But Netflix also invests more in original content than Apple does, too, and probably by a wide margin. I never quite understood why Apple got into content, it’s expensive and a risky investment, with no clear payoff. I suppose on some minor level you might view Apple’s decision to license its content–some of which is terrific and all of which is very well-made, two things that are no longer true at Netflix as it pursues a quantity over quality strategy–as being similar to what Microsoft is doing with Xbox: Perhaps this thing can be more profitable–or, just profitable–if it’s expanded to a wider audience.

Leo pointed out that Apple Music is small compared to Spotify. News+, I can’t see it being big, or Apple Card. Maybe Maps? Applecare? AppStore Revenue? Where does this second-biggest business come from, when a few years ago it was less than 10% of Apple’s business?

I’m not sure that Apple Music is all that small. It has more subscribers than Spotify in the US, and all of them are the more lucrative paying subscribers, where a big chunk of the Spotify user base is ad-supported. This is similar to the model Apple uses in other markets. But where the Mac is just 10-12 percent of the PC market, Apple Music is probably one-third to one-half the size of Spotify by users–these numbers are hard to find–but also more lucrative per-user. It’s hard to say.

But the relatively lackluster nature of most Apple services–Apple Arcade, Fitness+, and News+ have to be tiny from a usage perspective–are bundled the way they are in Apple One. Services like iCloud+ and, to a lesser degree, Apple Music and TV+, are probably doing OK, and are all high-quality. Apple’s customers are very loyal. And the big piece, again, is the App Store. It paid out $60 billion to developers in 2022, which means Apple made earned a nice chunk for itself too. Its infrastructure could run on a tiny percentage of the fees it collects, as we’ve learned in court cases. And it has an advertising model for placement as well.

The brilliance of Apple’s Services business is that it remonetizes existing customers in new ways, and it ways that happen more regularly. So someone will buy an iPhone every so often, and maybe an iPad, Mac, Apple Watch, or AirPods. Those are big. But Services are a monthly charge. And that smooths out things financially in a very nice way.

Best games of all time

spacecamel asks:

To continue your question from Windows Weekly, what are the top 5 video games of all time? I would go with Super Mario Bros., Doom, Minecraft, Fortnite, Madden.

This is one of those impossible questions in some ways. It’s not clear that a game like Pac-Man is in any way comparable to Call of Duty, for example. And like most, my experiences are limited. I played what I played, and I’ve missed out on some big titles over the years because of the choices I made with platforms. I have specific preferences, etc.

That said, I have ideas. And they can’t really be trimmed down to just five choices.

The original Spacewar!, which begat Space Raiders on the Atari 8-bit computers, both have to be in there.

Adventure, the text adventure, not the Atari 2600 game, which is probably more correctly named Colossal Cave Adventure. It’s gotta be on the list, and maybe alongside follow-ups like Zork. “You are in a maze of twisty little passages, all alike.” I mean, come on.

The early arcade classics like Pong, Breakout, Space Invaders, Asteroids, Centipede, and so on. Amazing. Pac-Man. Later arcade games like Zaxxon and Commando.

Graphic adventure games like King’s Quest and all the spinoffs, sequels, and similar games from Sierra and others. Maniac Mansion from Lucasfilm Games (LucasArts?). Classic. The Secret of Monkey Island. And Monkey Island 2.

Turrican and Turrican II are among my favorite side-scrollers of all time. I played them on the Amiga, but I think the were elsewhere too. Shadow of the Beast, but only the Amiga version. (And Beast 2 and 3, too.) Prince of Persia. Elite (or maybe it was Elite II.)

Arkanoid, a better version of Breakout. Lemmings. Many Cinemaware games, but especially Defender of the Crown and Wings, both of which I played repeatedly. Leaderboard Golf and Larry Bird vs. Dr. J: One on One. Pinball Illusions, so addictive. Archon, a fantasy chess game.

DOOM and DOOM II, obviously. Quake. Duke Nukem 3D. Unreal and Unreal Tournament. Half-Life and Half-Life 2. Medal of Honor, Call of Duty, and Call of Duty 2. And then the original Call of Duty: Modern Warfare series. Portal and Portal 2.

Halo, Halo 2, and Halo 3. The first Gears of War. Microsoft Flight Simulator. Tomb Raider has to be in there somewhere. The first Assassin’s Creed. Bioshock.

The first three Super Mario Bros. games and maybe Mario 64. Sonic the Hedgehog definitely. Minecraft.

I’m forgetting more than I remember, I bet. Especially in the third-person shooter area.