? Ask Paul: December 26 ⭐

Happy Friday and Happy Holidays! Here’s a short-ish edition of Ask Paul for this short week and extra-long weekend.

? The outlook for Outlook

Christian-Gaeng asks:

Since the Outlook team wants to relaunch the development of the new Outlook around AI, are there any indications that the classic Outlook will remain available beyond 2029? I can’t imagine that large corporations will switch to a new Outlook with AI at its core.

So many thoughts.

The first of which is that there is almost nothing I dislike more than Outlook in all its forms. But most especially the classic desktop app that some can’t bear parting ways with for some reason. The new Outlook is a sort of gift that keeps on giving if all you want is drama because there will always be some feature that one person finds indispensable that isn’t in the new app, and we can just complain about it endlessly. It fills the gap nicely between other dramas real (Crowdstrike) and imagined (Recall).

But the combination of the new Outlook and AI? Oh. Chef’s kiss. That’s the unholy nexus of Hell right there. It’s almost too perfect.

Outlook has thus far weathered two major shifts: Internet email, which arrived just as Microsoft released the first version of the app, and chat-based collaboration as popularized by Slack. In the latter case, Microsoft of course created Teams and has seen incredible success, but it also did that classic Microsoft thing and didn’t leave Outlook users behind by figuring out ways for them to stick with their tool of choice and interact with those who were on the new thing. Also, the whole “One Outlook” thing makes sense in that having a single codebase for add-ons across all app versions is obviously the right choice. (Cue next round of complaining about web apps.)

Quick side-trip: Thinking about this, it’s pretty clear that the technology behind Loop is/was the connective glue that makes all that work. And that in the pre-AI era, we were suffering through a similar volume of regular updates all tied to Teams. Pre-pandemic, during the pandemic, and then for the first few years after, it was all Teams all the time. From my perspective having to cover the news, AI/Copilot is just the Teams playbook, but amplified.

Obviously, Microsoft is going to integrate AI into the new Outlook. Right? I mean, this is obvious.

As obvious, Microsoft is not going to integrate AI into the classic Outlook. This can be viewed cynically, it’s so easy, but pragmatically (this will come up again and again, I bet) it makes sense. Classic Outlook is essentially if not literally deprecated, and it will get security updates, but it also has an end date. Though there is arguably some precedent for thinking otherwise if you’re familiar with the history of OneNote, Outlook is too big and too central to the whole Microsoft 365 productivity story to co-develop two versions of this app for the desktop. There can be only one.

Another side-trip: Complaining about the new Outlook to me is like complaining about the new Sonos app in that, yes, it’s terrible in so many ways, but we’re forgetting that the previous one was also terrible. And what is it we’re trying to save here? Some slightly less terrible garbage that’s built on a shaky legacy foundation? I mean, I get how attached one can be to tools. And it took me a long time to make similar changes, but … Outlook? It’s past time to Old Yeller this thing.

Sorry, I get distracted.

News of Microsoft planning to relaunch Outlook development to focus around AI comes via an internal memo viewed but not published by the Verge. So we have to just accept the few quotes they cherry-picked, and I will add the caveat that every time someone from Microsoft describes the addition of AI to whatever product as a “complete rewrite” or similar, my eyes roll back in my head so quite it almost physically hurts. Windows was described this way recently. I think we all understand no one is rewriting anything here.

But this is an internal memo, so the language is interesting, even if its author, Microsoft corporate vice president Gaurav Sareen, was perhaps anticipating (hoping?) that it would leak. We may not think about this all that often, but the other side of the coin we do complain about, Microsoft jamming AI down our collective throats in every way imaginable, is a more internal battle in which Microsoft executives and employees are being rated based on AI adoption. So in a way, this internal memo amounts to PR. This guy is trying to justify his position and role, and his product’s position in role in this new Microsoft. The message isn’t that AI matters, it’s that Outlook still matters. And it can keep mattering because it’s such an obvious place to integrate AI functionality.

Side-note: If you look Sareen up on LinkedIn, you will see that his tagline is “Engineering Leader | AI Enthusiast | Microsoft 365 Innovator.” Yes, I just threw up in my mouth a little bit. But also, that all feels so calculated. AI enthusiast? Dear God, man.

Anyway, Sareen claims to his team that the goal is to “reimagine Outlook from the ground up” and not just “bolt AI onto legacy experiences.” They will do this with Copilot, of course, as it will transition Outlook from “a set of tools into a partner that acts.” So he’s referring to agentic AI there, which is Microsoft’s current focus, which is what makes this make sense from a strategy perspective. For him, his job, and his team. And maybe the product, too, if anyone can really figure it out.

Sareen does write one thing I like, telling his employees to “let go of old ways of working.” This is good advice for anyone using Outlook, really, and especially the classic version. But OK, I know. There are features you really need, just can’t live without. Microsoft will eventually kill that app. And when it does, whenever it does, you can be sure that the new Outlook will not support some of those features.

As for the 2029 date, it’s not clear if Microsoft will extend that or stick to its guns as it (sort of) did, or is doing, with Windows 10. But just looking at the revision list at the top of this Outlook Blog post is instructive, suggesting that the schedule is, perhaps, fluid. And there are various milestones we’ll pass in which the new Outlook is recommended, becomes the default in more situations, and, in time does replace the classic Outlook. Each of those triggers a new round of angst, of course.

Ultimately, I’m not sure whether the details matter so much here as classic Outlook has to die sometime, and the sooner the better. One can make a much better case for retiring this 1990s-era mountain of slop than is possible for Windows 10, and they followed through on that threat. What maybe does matter is whether AI in Outlook will ever be off by default and opt-in, and easily ignored. Or whether Microsoft will simply do what it always does and promote this incessantly and/or just force it on us all. Were I betting money on this, I would go with the latter.

AI in Outlook is inevitable. Even those who hate AI should see the benefits of this even if they don’t want, it’s one of the correct places for AI, one of the end points where AI actually makes sense. Those who hate AI tend to be older. Those who cling to the classic Outlook tend to be older. And the overlap there is enormous, I bet. But if classic Outlook is still alive and kicking in 2029, it will be over 30 years old. And those who rose through the ranks during that era will be nearing retirement or just retired anyway.

I can’t really speak for the enterprise and other larger organizations that use Outlook. But the issue with AI, whether it’s “slapped on products” as Sareen says is often the case, or more fully integrated into a “rebuilt” version of the app, is just marketing. AI isn’t a thing. It’s a million little things. It’s what I would call features. So, it’s going to be all over Outlook. And I do think that most of those organizations will want that, and that Microsoft, for all its faults, is the right company to make it happen. The history there is clear, and its customers will expect, and get, all the data protections they require. Worrying about “AI in Outlook” is like worrying about spell-checking in Outlook. To me, at least.

I guess I could have answered this a lot more quickly by noting that Microsoft just rearchitected Outlook via that “One Outlook” initiative, and it will or will not rearchitect it again for AI, and that there’s no version of this story where classic Outlook limps into that new world. None at all.

? Changes

wright_is asks:

I was reading the What I Use thread by Markld on the forums. I have made some big changes this year, one being that I haven’t bought a renewal for Microsoft 365, instead I got a Black Friday deal on Proton and moved my private mail address there and started using some of their other services as well. I also spent several hours moving most of my accounts from using my Gmail and Outlook.com addresses to using my private address. There is still some work to do there, but 80% are probably switched now.

Nice. That’s difficult and time-consuming work and it’s the type of thing I keep wanting to do more of, until I try and I’m reminded of how difficult and time-consuming it is. I have started making a list of some related goals for 2026, one of which is finally consolidating my two Gmail/Workspace accounts. It’s daunting.

Have you tried Proton Lumo? It looks interesting, but seems like all other chatbots on first glance, but I hope that the privacy is as good as they claim.

Yes, and I think there’s a future in this sort of solution. Whether it’s Proton Lumo, Duck.ai, Brave Leo, or whatever else, we will see smaller language models and open source language models that are “good enough” and then actually rival the biggest and best models. And we will see services, paid and free, that let customers interact with the big LLMs in an anonymized way (like Apple does with ChatGPT in iOS). And the combination of these things will level the playing field, so to speak.

Some of the services I probably won’t use any time soon – I already get 1Password Family for free, because we use it at work, I have to use Microsoft Authenticator for my M365 accounts, so I tend to use MA for other OTP codes as well at the moment, and I haven’t yet had a need to try out the VPN service, for example.

Yes. Proton VPN routinely ends up at or near the top in any VPN comparison, but I only need this kind of thing infrequently. I did switch to Proton Authenticator this past year, but the Microsoft accounts just work with Microsoft Authenticator in unique ways, so that’s a blocker for me as well.

I haven’t switched to Proton Mail, but testing to see whether this might be viable is on my to-do list for 2026, as well.

I also switched from PiHole to NextDNS this year, as the PiHole PC (a Raspi 3b) was getting unstable and my wife was complaining about the Internet being down – it wasn’t, but, without DNS you can’t really get on the Internet these days.

I’ve stuck with NextDNS, which seems to work well, but I only enable it on my phones and iPads. (Note to self: Add testing this on PCs again to the to-do list.)

How is your move away from Big Tech services going? I’m guessing you will need to keep your Goolge One going for the business, or have you looked at moving that as well?

My goal isn’t to move off Big Tech services entirely. It’s to assess where doing so makes sense, make changes as needed, and to keep on that as things change. This is where the pragmatic bit I predicted would come up again is now coming up again. Which isn’t much of a prediction since I’ve been thinking about this a lot lately. I alluded to this in From the Editor’s Desk: Early Influences ⭐ because that came out of me going through all the year’s news and trying to pull out whatever trends and top stories and I realized that I was moving decidedly in this direction now. So there’s a coming post, probably to be called Pragmatic, that will focus more on that.

We’ll see what this looks like in article form, but in the notes I have now, I have written that “This is not a total rejection of Big Tech, but rather an open-eyed understanding that these companies will never value me as a customer and I need to aggressively look out for my own best interests.” I see myself “leaning, if not veering, away from Big Tech when it makes and towards what I think of as Little Tech.” And that the key to this kind of thing is to be pragmatic, as opposed to emotional or even unthinking. I’m not trying to make a statement. I want to do the right thing for myself.

Most of that is personal, I guess. But I do have this business, and it does have whatever costs. And these are things to keep evaluating as well. You mention Google Workspace, which is a bit expensive, I guess, since we have several users, each with some per-month cost. But if you think of value as a cross-section of cost and utility, Workspace lands in a good place in the sense that the cost and difficulty or moving to some alternative service combined with whatever loss of functionality makes leaving Workspace undesirable. Things can change, of course. But for now, I don’t see leaving Workspace. I am open to it if it needs to happen.

We live in this world as it is, and my Amazon, Apple, Google, and Microsoft accounts are mostly unavoidable. That’s true if only because we have to use smartphones. But even on PCs, we have this complex, interlocking series of Big Tech services and while they may be somewhat avoidable, they’re not completely avoidable. We do live in a golden age of alternatives, which may be one of those 2025 trends, we’ll see. But I can use something like Brave Search most of the time and it’s fine, but when I don’t get what I need, I do find myself going to google.com and searching again from there. That will be the case with Lumo. It may be the case with LibreOffice, where you get some document that’s formatted in some way that only real Microsoft Office works, etc. It kind of never ends.

I could go on, but I do have this partially written article and all the other 2025 wrap-up stuff in progress, so there will be more in the next several days. Long story short, I have been wandering in the wilderness for years, and this past year was a clarifying moment for me in which I am now moving forward with purpose. And I will lessen my exposure to Big Tech where and when possible, not to send a message but to improve my life. I’m paying attention now.

Edit: I just got an email from Proton about their charity fundraiser, I find that a really good idea.

Agreed. For those not on Proton, the company is hosting a fundraiser in which it is raffling off 10 Proton Lifetime accounts. The raffle tickets (e-tickets?) are $10 each, you can buy as many as you want, and it runs through January 5. If you’re a Proton member, there are always ways to get 1 to 50 free entries.

? All I want for Christmas is that thing I already bought three months ago

train_wreck asks:

Short question: How’s Christmas been?

We came back to Pennsylvania specifically for the three holidays—Thanksgiving, Christmas, and New Year’s Eve—because this is family time and a good time to catch up with friends, and not for the weather that may or may not ruin this coming weekend. So the kids are here now, which is terrific. Mark (our son) and I stayed up stupid late last night, perhaps drank a bit too much, and enjoyed watching Patriots and other Boston Sports-related videos on YouTube. And these times are priceless.

I’m not sure whether it makes sense to write about any of this, but my wife and I agreed to really minimalize the gift giving between the two of us this year. So we opted out of several small things that we would otherwise put in the stockings, and we set a limit of roughly $100 to $200. I got her a CarTablet Pro Max Apple CarPlay/Android Auto display for the car. And she got me a pair of Anker Soundcore Sleep A30 ANC sleep earbuds, which I used last night and are terrific. (She also got me a toaster because the toasters we have in both places heat up and harden bread but don’t actually toast.)

ℹ️ The intel on Intel

train_wreck asks:

Lip-Bu Tan has gotten some press recently about how he was the “dealmaker” who greased Trump’s wheels to get their big government investment. Do you think this is an accurate portrayal?

No.

I do think that Tan played Tr#mp like a fiddle, but that’s a low bar: Every time he meets anybody, they turn into the greatest human being that’s ever existed. So it’s like stealing candy from a baby. A big, fat, corpulent baby who is intellectually and morally challenged and has whatever other problems but is, above all else, just a baby.

Tan doesn’t strike me as particularly skilled at anything that will benefit Intel and he’s not well-spoken, so he doesn’t have a great on-stage presence. Pat Gelsinger almost single-handedly created the CHIPS Act out of some incredible combination of PR and grit, and it was only because the U.S. government never actually paid them a cent of the $8.9 billion it promised that the more recent “investment” occurred under Tan. Can be “credit” him for that? I mean, I guess so in that he went through the whole supplication routine. But I feel like Gelsinger could have orchestrated it pretty handily had the Intel board just given him another year. In fact, he may have gotten better terms given how freaking stupid our president is.

What I do give Tan credit for is taking the Gelsinger turnaround plan and making one crucial addition: Massive cost-cutting. Intel is now aggressively dropping businesses that aren’t paying off, and that should have happened more immediately. And no one likes layoffs, and Intel has had a ton of them this past year (reference: Me going over the year’s news for the wrap-ups), but this, too, is very necessary for a company that had gotten too big and too directionless.

Sometimes that requires someone from the outside, I guess. Gelsinger worked at Intel for decades, but Tan was just a board member. If it weren’t for his new job, none of us would have ever heard of this guy. He’s also on the board for UC Berkeley. Perhaps he could run that organization next. 🙂

Happy Holidays!