Ask Paul: January 20 (Premium)

Image source: The New York Times

Happy Friday, and welcome to another set of great reader questions to kick off the weekend. This article was NOT written by AI, I promise. 🙂

Overreactions

leoaw asks:

In the past few days I’ve had a series of emails from stores that used to offer various member discounts now get rid of them. Best Buy is switching its Rewards to only people with their credit card. Amazon is getting rid of its Smile donation program. A gas station rewards program I was in just sent an email today that they’re changing their program, too. Is there an external regulation driving this (I did receive a bunch of “our privacy policy has changed” emails at the beginning of the month) or has the industry collective determined they weren’t getting their return on investment in the programs?

I feel like this might be related to the economy and that each of these companies has finally started taking a hard look at their business, determined which bits are costing them too much, and have started scaling back. Brad and I talked about this, sort of, on today’s First Ring Daily, this notion that Big Tech (in the case of our discussion) was only too ready to accept what should have immediately been seen as a very temporary surge in pandemic-related revenues for what it was. But instead, they reacted like Lotto winners, and over-hiring was just one side-effect. I’d like to believe that these firms are led by adults who understand how these things work. But the pandemic proved otherwise.

To your specific point, I guess those examples aren’t as dramatic taken one by one. But as a consumer and an individual, you can literally feel the belt tightening at all of the companies whose products and services you use. And they are collectively having an impact at the individual level. I saw this at Dunkin’ Donuts, for example, and the change they made to their rewards program a few months back was so confusing and so unfair compared to the previous system that I literally stopped going there. This is what happens when companies overreact: they did so during the pandemic and now they’re doing so in the opposite direction in the post-whatever-recession. And it makes them look less sophisticated and in control. It’s like someone spilled money into the air and they’re running around, bumping into each other trying to catch the bills before they hit the floor.

Ultimately, the only power we have as consumers is to take our business elsewhere. And while we may be looking at reduced benefits no matter where we turn, depending on the market or business, there may be some options that aren’t taking these major steps back in customer service. Or at least some that aren’t as bad. And we vote by changing.

I’m going to write about Google’s overreaction, specifically, later today or tomorrow depending on my schedule goes.

Mac Mini vs. Mac Studio

helix2301 asks:

Paul, what are your feelings about the new mac mini vs. Mac studio?

I’m not an expert in Mac hardware configurations, but I was intrigued by an article I read describing the areas in which the new M2-based Mac Mini still doesn’t compare well to the last Intel-based version, which is now discontinued. The Apple Silicon chips have been incredible for Apple, but their integrated design also gives up some capabilities, too, around such things as discrete graphics, the number of supported external displays (on all but the very much expensive and most recent chip), and expansion ports. And that stuff can be important depending on the audience.

Leo and macbreak weekly crew all pointed out that the upper-end mac mini is almost as powerful as the low-end mac studio for a lower price.

Well, Apple is no stranger to overlap: I wrote about the weird functional overlap in the current iPad lineup recently, for example. Because of the odd, slow way that Apple updates its products, there are these gaps in time where a new but lower-end product might be better in some ways than an older, but the higher-end product. I think that’s what we’re seeing here: the new Mac Mini is based on the new M2 and M2 Pro chipsets, while the Mac Studio, which was not revved, is based on the previous-gen M1 Max and M1 Ultra chipsets. There are other differences, of course, but one assumes we’ll see M2-based Mac Studios at some point that will put this argument to rest. Honestly, the best news about the new Mac Mini is that they brought back the $599 starting price. Regardless, Mac Studio is incredibly expensive.

Still no mac pro yet?

No, but the indication (from Mark Gurman, mostly) is that it’s coming and will not be as advanced as Apple had planned because they ran into chipset scalability issues. As a result, it will just use the same enclosure as before.

This is not my area of expertise (assuming I even have one). But that’s sort of how I see it from the outside.

Windows 11 futures

helix2301 also asks:

The joke around the office yesterday was, now that we have an EOL date for Windows 10 I guess it is not the last version of windows. What are your thoughts on 11 will it become the new Windows 8 or Vista because adoption will be slow because of its great predecessors like Windows 7 or XP?

This will all depend on what happens next: do we get a Windows 12 in 2024? And if so, does this signal the return to a 3-year release cycle? Or does the presumed “AI-ness” of Windows 12 signal another Windows 10-like schedule where the same product stays in the market for several years?

We don’t know the answers to any of that, but I think it’s fair to compare the relationship of Windows 11 to Windows 10 as being similar to that of Windows 8.1x to Windows 8.0: they’re the same thing internally, but they’re going in a different direction with UI/UX and now all future updates will only be applied to the newer thing.

What I’m always sort of curious about is whether any of these things will stick around long enough to achieve massive scale. When Windows XP was kept in-market artificially for much longer than originally intended—thanks, Longhorn—it became the best-selling/most-used version of Windows ever, and the assumption was that that record would never be beaten. But then Windows 7 surpassed it for similar reasons: its successor, Windows 8, was hated and ignored and Microsoft had to extend support for that system for an extra three years, meaning that it had an even longer active lifecycle than XP and then became the best-selling/most-used version of Windows ever. But thanks in part to a bigger market, the Windows 8 hangover, and it too being kept in the market longer than had been the case to that point, Windows 10 is now the best-selling/most-used version of Windows ever. I can’t say that is as well-liked or respected as Windows 7, I guess. But numbers are numbers.

But that was the goal: to get as much of the user base on one version of Windows as possible to ease updating and respond more quickly to threats. Now, we seem to be stepping back into the previous world of three-year active lifecycles. But note that this hasn’t happened yet. We’ll see.

Xbox

helix2301 also asks:

My daughter wants a new Xbox she has a few years old xbox one s from years ago is it worth upgrading she is not getting an X she does play Fortnite.

Yes. You will be very pleased with the Xbox Series S, and it can be quite inexpensive. I’ve been using one for probably about a year now, and it works great.

OpenAI, open skies

hastin asks:

Paul: Do you see any of Microsoft’s big AI push being moved into Windows (the OS product) at all?

Yes.

Personally, I’d love to see Windows more self-heal, understand bad actors (like phone scammers opening command line), and use AI to benefit the platform and its stability – not just power the creation apps inside of it.

Microsoft would probably argue that there is AI all over Windows today already, but yes, I do think that AI permeates the entire Microsoft stack in the coming years, and that the presumed Windows 12 will heavily push those advances. There is literally no end to the ways in which AI can be used to improve Windows features, and just trying to describe them can quickly drift into what feels like science fiction speculation. But I think there’s a good guide to how this can be presented and marketed in Google, which today promotes its products and services as “helpful.” For example, the most recent Pixel lineup was announced as being designed “so the helpfulness and intelligence of Google can adapt to you in a non-intrusive way.” This is generally how I see Microsoft improving all of its products, including Windows. Or at least trying. Clippy done right, perhaps. (Where Clippy was an odd combination of intrusive and unhelpful.)

It’s hard to get specific here. But given Microsoft’s attraction to telemetry data, it’s not hard to imagine it targeting those actions that some large percentage of people try to accomplish in Windows and figuring out helpful ways to make them successful. This ranges from the smallest tasks on up.

It’s going to be an interesting couple of years.

Apple Books auto-narration

spacecamel asks:

Did you see the announcement from Apple on integrating digital narration into their Apple Books?  I think the samples they showed are impressive and closer to real voices than what Amazon or Google have shown. 

Yes, but I’ve not tried it yet. I’ve been meaning to do so. This is an interesting capability, and it really undercuts companies like Amazon that offer both ebooks and audiobooks and charges you twice to have both. Maybe that’s a good thing, and “revenge from beyond the grave” on Apple’s part for Amazon (mostly) prevailing over it in that iBooks antitrust case in which Apple colluded with publishers to raise ebook prices.

I need to take a look at this. Digital voices (as in Edge read-aloud) have gotten a lot better in recent years.

AI and content creation

will asks:

In reading your recent post on the AI era, I was thinking about what this looks like for the future of content and information we get going forward, and I am a little concerned. As you know with today’s version of AI tools you can create photos and images that look and feel as if they were created.  The New York times had a recent article on a 70’s Tron movie that never existed, but the images look as if they were taken from the movie.

Yes, I read that article with much interest, but it’s worth pointing out that AI didn’t make the movie, it just did what WALL-E does and created images based on some inputs. They’re impressive, for sure.

How much harder would be to have a full custom movie created based on information someone feeds into it?   My point is that it would be easy for the AI generated content to be taken as if they were real, or really happened. Photos of events that did not happen or happened in a different way.  Same goes for writing and the information we read in the news or online. Articles and information could be created in seconds that could sound convincing, but it was all generated by AI.  Future news articles could all be AI written, real-time, as events happen. Curious your thoughts on us slowly building SkyNet one photo at a time?

My wife, who is also a writer, was asking about this topic this morning. She was curious if I had ever used any of these tools, which I haven’t, and whether using them made sense. What’s going to happen, I think, is that credible writers and other content creators will start using these tools but heavily curate/edit the results, while the less credible will just present AI-created works as-is. (You can imagine the high school student book report or whatever.) And her follow-up to that was to wonder whether it would be faster or more efficient to just write an article on whatever topic or to have AI generate it and then fact-check and edit it herself. And … it depends. More to the point, the latter will likely become more viable over time.

One of the more interesting aspects to this is when—not whether—AI can successfully replicate the voice of a writer. For example, I write in a certain style, with long sentences and lots of commas and em-dashes to separate parts of those sentences. And I have my own viewpoints and would like to believe that I take on various subjects from an angle that can sometimes surprise people. (“Think differently.”) Can AI duplicate that? Probably. Maybe. But in the future certainly.

Does that mean my future is sitting on a beach sipping a drink while AI writes my words for me? No, but some will no doubt do that, if not now then someday. I will start looking into how or if AI can help with writing, but I’m not so lazy to think it can do the heavy lifting. I feel like my best ideas come to me in moments in which my mind is wandering or somewhere else or, better still, while talking through something where I suddenly find myself heading down some path. Whatever the process, I’m not sure AI will ever duplicate that. But the question is whether it can augment that, or answer questions in a way that current tools do not. This requires investigation. And perhaps time for these tools to improve.

Put a little more simply, yes, news articles will be written by AI, there’s no way around that. The tricky bit is the editorial angle, the perspective. An AI can have a planet’s worth of information at its disposal, but will it have perspective? Will it understand humans and how we react to things? Maybe someday. But are we there yet? No one knows, I guess. I don’t.

There is a human connection that drives what I do, at least. When I was first hired at what became Windows IT Pro—it was literally still called Windows NT Magazine at the time—I was told that my writing was too personal, with too much “I” and “me.” This was only natural to me, and it was interesting to watch this editorial-heavy organization slowly realize that it was that sort of thing that made my writing interesting to people, that technical writing didn’t always have to be like dry reference material. It was sort of the blogification of technical writing, I guess. They didn’t get it at first.

Anyway, I am curious to see whether AI will ever duplicate that, not just the style, but literally the ability to go where my brain might go. It seems possible, right? But I’m not interested in letting some “thing” write for me. I am curious about how it might help, in the same way that something like Google Search can help me mind facts or references so quickly now, in a way that was impossible 25-30 years ago.

Sleep problems

simont asks:

What do you think of this problem with sleep for Windows laptops? Justified or hard technical problem?

Microsoft is Forcing me to Buy MacBooks – Windows Modern Standby and Microsoft’s Answer To LTT (Windows Modern Standby being the story so far if you’re not aware of the issue.

As someone with sleep apnea, I can relate.

But seriously, both videos are great examples of hyperbolic click-bait: this problem, which never seemed to impact me once ever over the past year despite reviewing so many laptops, is so bad that the first guy is forced to buy a MacBook. Sigh.

Long story short, Windows supports various power management modes and can retain an efficient “heartbeat”-style connection to the Internet when the PC is sleeping so that things are up-to-date when you wake it back up. The canonical example is email: you wake up the PC, switch over to your mail app, and all your new mail is sitting there waiting for you. But this also impacts things like Windows Update, which can “drip” download updates and even reboot to install them when the PC is asleep. Etc.

But the notion that PCs are somehow hot-bagging regularly here in 2023 is, to me, almost fanciful. And unlike most people, I base that observation on having used not just a lot of PCs but a lot of really modern/brand-new PCs. I just don’t see this happening. But in the second video above, the two hosts are acting as if this is the dirty little secret of the PC world, that this happens all the time to everyone. This is a conspiracy-level exaggeration if not nonsense. Not because it doesn’t happen at all, but because the way we’re wired is to remember the extreme things. That second video waltzes through that exact topic when the guy on the left catches himself. It’s not “every time” he grabs the laptop, it “feels” like it, “probably not half” the time. Then it is all the time again. Because perception.

(Meanwhile, none of these same people or any other reviewers to my knowledge have ever experienced the easily replicated USB/TB dock/hub issues I’ve seen again and again with PCs based on 12th Gen Intel Core chipsets. This is also frustrating.)

The complaint seems too technical for most people: Microsoft is removing S3 sleep from Windows, “forcing” users to use S0. There’s a good article on Petri that explains the differences, but long story short, sleep works great in my experience. Most of the PCs I used this past year turned on instantly the second I opened the lid most of the time. I never saw hot-bagging, and certainly would have mentioned it if I had. I did sometimes have to press a power button. Oh, the horror.

Microsoft’s response is that it is moving away from S3 sleep because sleep states are controlled by a PC’s firmware, and that’s the responsibility of PC makers, not Windows. S0 sleep gives Microsoft more control over gives Microsoft more control—no worries if you’re not up on the latest firmware updates—and sleep is thus more reliable. Seems reasonable.

It’s also worth pointing out that the Surfacegate issues that Microsoft experienced in 2015 with the Surface Book 1 and Surface Pro 4 were very much about this exact issue. Those PCs had 6th Gen Intel Core chips, so we’ve had at least 6 new generations of Intel chipsets since then. (Was there a 9th Gen? I don’t believe so.) And these problems allegedly aren’t just on Intel-based PCs.

Anyway. Here’s my take. Bring your power cable with you. And if you’re flying, power down the laptop before you leave the house. Or, don’t worry about it because most planes have power now anyway. I can’t recall a flight in the past year that didn’t. And if you do experience hot-bagging even once, I feel like you’ll be more vigilant. But overall, this seems like too much fury for what’s happening. It’s exaggerated, and there’s a lot of head-nodding. And it just rubs me the wrong way.  There are real things to complain about with Windows.

VPNs

GertRoux asks:

Do you use a VPN? If so, which one?

Not at home, but I do use ExpressVPN sometimes when we’re in Mexico or abroad otherwise.

NextDNS

JustMe asks:

A recent article I saw discussing running a PiHole in Linnode (cloud-based) got me to wondering if you were still using NextDNS and if your impressions were still the same of it.  Have you considered other solutions since you settled on NextDNS?

This is interesting timing as I was just discussing how that experiment went with my wife using it on her phone. Based on that success, I’m going to be enabling NextDNS on the router in the next week or so and make sure it doesn’t negatively impact things like Sonos or Apple TV. Assuming that works out well, the transition will be complete. Meaning, I’ll see it on my PCs as well (at home). Right now, I’ve only enabled it device-by-device on all the phones, my iPad, and some PCs for testing.

NextDNS works well enough that I’ve never considered trying anything else.

Browser extensions

jchampeau asks:

What browser extensions do you use?  Do you have a good one for dealing with cookie notices?

I use Brave, so I need fewer extensions than I use on Chrome, Edge, or other browsers, and Brave recently implemented a cookie notice bypass, which I enable when prompted.

As for the extensions, on Brave, I have Bitwarden (password management), Dark Reader, Google Translate, Grammarly, Momentum (new tab dashboard), Save to Podcast, and Simplify Twitter Web UI.

On Chrome/Edge, I add more: Adblock Plus, Disable HTML 5 Autoplay (Reloaded), Postlight Reader (Chrome only), and uBlock Origin.

Eero successes

CrownSeven asks:

How are the eero’s working out for you so far?

This system has been a game-changer, and I have been coming around to the notion that most of the connectivity issues we’ve experienced over the past 5 years (and there weren’t that many) were in fact usually caused by the Google Wifi system and not RCN/Astound. The Eero has been rock-solid the entire time, the speeds are higher, and it reaches further into the house with fewer nodes. I could not be happier. I wish that I had upgraded earlier in some ways, but the Wi-Fi 6E-based system we did get should last well into the future.