John Voorhees tried the new ChatGPT Work sign-in feature and blogged the details at MacStories:
If you can get logged in, having an agent collect and analyze things like analytics data is fantastic. However, getting past the initial login screen is still too frustrating.
It is remarkable this works at all. OpenAI has all the pieces in place... But I hope they can consolidate Codex, Work, and Chat in a way that isn't so confusing between local AI and the cloud.
Reading about the Meta settlement. I'm surprised that the changes are actually pretty significant. Nighttime block, muted notifications during school, and no like counts. Of course they could do more, but it's good progress.
Coffee shop this morning has Dolly Parton on repeat. 🎶
I watched a bunch of videos from this week's World Humanoid Robot Games in China. This segment on YouTube is an okay place to start.
These robots are advancing more quickly than I was expecting. I've blogged a few times that humanoid robots are a bad idea, and now I'm really convinced of the danger.
There will be helpful robots that fold your clothes. There will also be robots working in factories. And at the other extreme, there will be military strike teams of a dozen robots led by a single human. I'm not an AI doomer — quite the opposite! — but it's easy to imagine the worst case here.
I think small, special-purpose robots would've been a better path. R2-D2, not C-3PO. We imbue human traits in anything that looks humanoid, blurring our perception of the line between machine and sentient being. When these humanoid robots crashed out in China, I almost felt for them, even though they are designed purely for movement with no consciousness.
The future is going to be wild, and it appears too late to change the trajectory of robotics except with small, incremental corrections.
It was only a few weeks ago that I rewrote all my blog links to X to use XCancel.com instead. Now because of a cease and desist, those links are broken.
Micro.blog also archives everything you link to, but for some reason it doesn't appear to work with XCancel. More digging required.
Dolly Parton was one of a kind. We still watch 9 to 5 and Steel Magnolias sometimes. The “kindness” billboards featuring Dolly are a perfect tribute. Rest in peace.
We've been catching up on Apple TV shows. Watched: Lucky. I wish they'd make a season "0" that is all heist, but I liked how it wrapped up. 📺
Holy moly. A maxed out Mac Studio is $18k. If you wanted a nice Mac with lots of RAM, you missed the window. 💰
I'm still amazed that AI prompts like this work. We've barely scratched the surface of what is possible.
I take storing photos very seriously. Even with the EU outage this morning, there won't be data loss. We put photos in a couple places in case something like this happens. For Mac users, I also love the Backup tab in our Micro.blog app. Full blog posts and photos backup in a single local file.
Micro.blog hosting is spread across Linode, AWS, and Hetzner. I've posted a couple of times asking how Hetzner can be so inexpensive... Today I found out. Major outage with their S3-compatible storage.
Looking over the specs for the new Mac mini and Mac Studio. Not sure the Mac mini is a good box to put in the corner of your house running AI anymore. $900 base price. Would've been cool to have a stripped down MacBook Neo as a low-cost mini. I pay $14/month right now for my little OpenClaw server.
Maybe one day we'll have a better way to verify humans on the web. CAPTCHAs are too user-hostile (and will become ineffective). Scanning eyeballs with an orb is too dystopian. Requiring a credit card or drivers license is clunky and risks privacy. Dunno.
I got access to W Social, so I can now test cross-posting from Micro.blog and tweak any Bluesky-only assumptions in my code. But the verification process is pretty complicated. QR codes, mobile app, scanning passports! This is a lot of friction.
Years ago, the principal at my kids' school would always remind students that when given the choice between being right and being kind, choose kind. This was a line in the book Wonder, adapted by author R.J. Palacio from a similar phrase by self-help author Wayne Dyer.
Being right matters, but it's not always the most important thing. Some people are so sure they're right that they justify treating people harshly. Worse, they might still be wrong! Let me illustrate this with a table:
| Right | Kind | |
|---|---|---|
| ✅ | ✅ | 🙂 |
| ❌ | ✅ | 🙂 |
| ✅ | ❌ | 🙁 |
| ❌ | ❌ | 😡 |
The best thing about "choose kind" is that it protects you from being wrong and an asshole at the same time.
AI coding can be so thorough that it creates a mountain of code to review and maintain, even for simple features. I'm still trying to find the right way to balance this superpower with my pre-AI instinct to write the least code possible. It's about deciding which edge cases matter.
Finally took another look at rewriting my Libby sync to account for changes in the Libby app. Early testing works, so I'll deploy it tonight. It'll be nice to have my book loans in Micro.blog again.
Mark Gurman has talked to people who have used the new foldable iPhone:
People who have tried the Apple model like how it fits in a pocket, the feel and durability of the advanced hinge mechanism, and the new iPad-like app layouts that take advantage of the larger internal display. It’s going to be an ultra-premium device that provides a halo for the full Apple lineup and will be a marketing engine to get people into the company’s retail stores.
At $2k, it's just too premium for me. I still think what I'd really want is a foldable iPad mini with no outer screen.
Wind turbine blade going up I-35. I'm always amazed at these, seeing them up close.
Reading a lot of replies on Mastodon, Bluesky, and even some on X to take the state of things. My blog post from last month is relevant this week. I don't expect everyone on Mastodon to agree, but I hope those who care about safety will at least consider it.
Anil Dash in a follow-up reply on Bluesky after blogging about how to move away from Big AI:
People tried to shame folks out of using Instagram, out of using Uber, out of getting into crypto. How’s that going? What if, instead, we had focused on providing credible, pro-social community-based alternatives that addressed the things people thought they would get from those platforms?
He's right that we can learn from the success of podcasting too.
Something I've been asking myself lately: what new things are possible when AI costs drop? I've been improving Micro.blog's Logs page, and this morning updated it to rewrite confusing Hugo messages in more friendly language. It even localizes if your account isn't set to English.
Love this. IndieWeb is punk:
Nobody’s waiting for permission to have a website. You don’t need a “content strategy.” You don’t need a niche. You need a place that’s yours, and the nerve to hit publish.
🎸
About halfway through Ride or Die season 1 and enjoying it. Hannah Waddingham and Octavia Spencer are great together. 📺
Skimming through some junk mail, always looking for things to unsubscribe from, and this is an actual email subject line from LinkedIn: “Barack Obama, President of the United States of America, is popular in your network”. No kidding! Algorithms losing the plot. 🤪
Federico Viticci and John Voorhees have posted an explanation about returning to X:
MacStories is precious to us. It’s not just how we earn a living; it’s one of a small number of independent websites that still cover apps, Apple, and a growing list of topics, including videogame hardware and the automation and productivity side of AI. Readers shouldn’t have to think or care about the business side of MacStories, but we have to, which is why we returned to X.
I’ve seen some people post about unfollowing Federico or cancelling their Club MacStories memberships. Is that really what we want, to withdraw support from a long-time blogger who has had such a positive impact on the indie Mac and iOS ecosystem? We undermine our own cause if we do that, potentially isolating people who we otherwise have so much in common with.
Instead, let's ask why this happened. I'm concerned less about Federico's choice and more that it's a signal of new growth on X, especially around AI. I'm increasingly confident that's the case, just based on keeping an eye on X this year. There's a timely thread by Mike Masnick on Bluesky that applies here:
Multiple people I know have told me that they love the idea of Bluesky, and want it to succeed, but have abandoned it for X because the use agentic tools in their work and find them incredibly useful, and feel that any mention of their usage here leads to hate and ridicule.
Driving people away from open platforms, so that they feel they have no choice but to rejoin X, will only make things worse for the web. This same "you're not welcome here" story has played out on Mastodon too, most famously with Wil Wheaton. Our communities should try to resist divisiveness or we will be splintered into smaller and smaller groups with even less influence, at worst becoming insular or even hostile to new users.
Please keep in mind the big picture. To Federico, if we were to chat about this, I'd encourage him to mirror all his X posts to an indie microblog or Bluesky. That way he can connect with folks on X while avoiding exclusive support of a silo and, indirectly, its leadership.
The social web is at another inflection point. We can't count on another wave of X users leaving for Mastodon and Bluesky. If we want to grow the open web, we're going to have to earn it.
While working on a new blog post, I asked AI a question about my older blog posts and it showed me this one from 2016: Blog when you disagree. I still think this is a powerful aspect of blogging. Posting to your own blog is inherently independent. You post for yourself instead of for a community.
Reading up on AT Proto Spaces. (Wait, when did we start spelling it Atproto?) It's private-ish, not encrypted, and kept out of the relay... I think this is going to be a big deal.
Apple responds to OpenAI's motion to dismiss. Yawn. I think when OpenAI shows their new device later this year and it doesn't resemble a HomePod at all, this lawsuit is going to feel pointless.
Watching video from WordCamp, Matt Mullenweg mentions they should be putting WordPress profiles on their badges:
Why are we putting like Twitter handles and other stuff. We are the WordPress people. We should be... you know our WordPress profiles and our blog URLs as the primary thing we're promoting, not these random other walled garden social networks, right?
I didn't realize that they owned w.org too. See also: "hello my URL is" name tags.
I don't know if this has gotten worse with Liquid Glass, but I'm noticing that simple context popup menus are just too slow on iOS because of the fancy animation. I'm tempted to write my own that opens immediately when speed matters.
Continuing to improve the Logs page in Micro.blog. I think the color for different tasks helps. Still a little nerdy, but it's a very useful page to understand what Micro.blog is doing.
Finally ran out. But then I remembered that the 5.3 credits are tracked separately. I had also forgotten just how fast 5.3 Spark is! For simple changes, it's not bad.
Did not finish: Water Moon by Samantha Sotto Yambao. I loved the idea and it starts well, but 25% in the writing didn't quite work for me. Decided to put it down, there's too much else to read. 📚
Listening to Fireflies by U2 and Martin Garrix. Can't wait for the new album. 🎶
Twenty years ago I participated in NaNoWriMo. I was just remembering how I had a printed book that was all synonyms and antonyms. Handy when I got stumped on just the right word to use.
That's how I think of editing with AI, except instead of words it's phrases. AI is good at telling me when I've written something poorly, something that would be improved by just changing a short phrase here or there. I treat it the same way I'd treat advice from anyone: if a friend told me this sentence was confusing, would I change it? Sometimes yes, sometimes no.
I don't bother with that for most of my blog posts. It's usually for the longer posts — where I'm trying to make a nuanced point and something feels off, or where the topic might blow up if it's misinterpreted — that I'll send the draft through ChatGPT for a sanity check. For some posts, in the prompt I will explicitly say "major grammar problems only", to avoid getting distracted away from my own writing.
Sometimes I won't make any changes. I just want some reassurance that my blog post isn't all wrong.
The value of getting help from AI is that because it's a synthesis of all the world's information, it tends to pull you toward accepted narratives. The risk is that it will water down your own unique ideas. Overusing AI for editing will produce a kind of sameness as everything else on the web. Taken too far, it becomes slop. Lots of words that don't really mean anything.
I mention all of this in part to be transparent, in part because it's relevant in the context of watermarking, and in part because I think I could still become a better writer. Even after blogging for over 20 years! I love to write and to blog. AI can't change that.
New trailer out for Wildwood. This looks extraordinary. Hope it's even half as good as it looks.
Micro.blog book pages are looking pretty good.
Seems like we might be settling in for a 5% Apple tax across the board, globally, based on the US court and EU compromise. I don't love it. Linking to external payments should give Apple 0% and nothing more.
Related, I saw this scare sheet yesterday while managing my account in Disney+.
I've never run out of Codex tokens before, so it's always felt essentially unlimited. At 4% remaining now with a day left! Uh oh. I don’t think my usage has changed much except that I’m running with /fast and Ultra more often.
Catching up on news from D23. Cartoon Brew has a rundown of some of the theme park changes coming up. The first Lion King ride will be in Disneyland Paris... A little surprised they never built that in Animal Kingdom.
My post earlier about X didn't resonate exactly as I had intended. Let me zoom out.
Let's think of proponents of the social web as a coalition. People actively choose to be on the fediverse, atmosphere, and (though it's a dated term) blogosphere. There are people who:
These are all different things. Which of these matters most to you will shape your perspective.
I believe that blogging is the best way to contribute to the web. Because of this, many other things follow. For example, it’s better to post to your blog and X (or your blog and LinkedIn, or your blog and Instagram) than it is to post to Mastodon alone.
We spend so much time on this social network vs. that network that we risk missing the big picture. Maybe the social web has gotten too tribal and in turn too far away from principles.
Pretty significant design changes coming to Mastodon for version 5.0. I don't use Mastodon often (my blog posts get to the fediverse from Micro.blog) but these seem like good changes.
Got let in to Bluesky's Attie. I'm impressed with it so far. I wonder if this will end up being a paid product. Can't be cheap to run.
I had missed that Automattic's renewed push into education comes with a free .blog or .art domain name. It's just for one year, but still great to get students thinking about owning their writing. I wish I could offer that.
When I blogged a couple weeks ago that X seems to be growing again, I think some people didn't believe me. Yesterday, Federico Viticci returned to X:
Over the last two years, I've fundamentally changed as a creator. I'm still a writer and podcaster, but I guess I'm also a new kind of "builder". Today, I need to find new readers, listeners, and developers, wherever they may be.
I have a lot of respect for Federico. If he feels he needs to connect with folks on X, that's his call. I'm sure he's going to keep blogging and podcasting too!
Let me be clear: I'm not going to post on X. But I'm not blind either. Part of running Micro.blog is paying attention when things change.
Looks like we're getting the camera AirPods soon. I'm curious how Apple is going to frame the privacy for these considering all the Meta glasses backlash. Even if the image never leaves your phone, will people trust it?
John Gruber follows up on a number of issues around watermarking. On what to even do with detection tools:
There cannot exist a single detector for all LLMs. And even if you find out that it says it’s a match, that an email or blog post or Slack message was very likely generated by, say, Claude, what are you going to do? March into your colleague’s office and tell them you caught them?
Watermarking critiques are compelling because there are multiple points against it. Worse quality (maybe), false positives with quoted or lightly edited text (likely), centralized control of keys (yes).
It's hard in the moment to know when to stand up to bullies and when to let it go. Glad to see Disney pushing back against Trump's FCC.
We just sent a new Micro Monday email update out! Still getting our bearings again with email. I also built a new system to manage and send the emails. Learning a lot, as always. 🤞