FeedCity

Apple says they’re planning to lock down Full Disk Access even further:

Going forward, we will introduce additional controls to ensure that users who genuinely wish to grant an app this extraordinary level of access can only do so with very explicit user action.

Sounds complicated / annoying.

Amazon has a long blog post about data centers that no one is going to read in full. Here’s one part:

…this build out is so important geopolitically that there are widespread reports of various countries intentionally seeding misinformation in the U.S. about data centers to trick us into slowing down.

It wouldn’t surprise me if this was true. But Amazon’s blog post isn’t going to convince anyone who is on the fence. AI, data centers, and humanoid robots are very polarizing and will continue to be. We get stuck talking about what we’re against and the divide widens.

I like the way AI is evolving, with Mush from Meta and 00s from OpenAI. Even the main programs are starting to provide real world services that the providers should be providing. They could possibly take over commerce the way Amazon has taken over shopping. There's room to improve in everything about the way we do business. And it seems like they've been researching this and experimenting (of course) and this is what they've come up with. Haven't tried them yet, but it's a good direction to go. It really should be Apple and Google doing this, btw. Also ChatGPT can help you understand your credit score. That's something I'd seriously like to ask some questions about.

There’s a lot of chatter about moving away from cross-platform frameworks, back to separate native code for each platform. I’m not convinced that is always a good idea. Creating apps is relatively easy now, but maintenance is still hard. For our 13 apps, we have 8 code bases. Overwhelming already.

Now that it’s October (!) I’m getting excited for travel later this month. Going to be near Seattle for a bit, and then… Can’t believe this: SEA → CHI → NYC on Amtrak. Just spent a little time reviewing plans and handing them to Dots to see if it can help keep me organized. 🚂

After running the first Bloggercon at Harvard in 2003 a famous tech company had a meetup to talk about the same stuff at a private invite-only session in the Bay Area. The invitations went to all the discussion leaders at Bloggercon. People I chose. I was not invited. Only o...

Today there was a rainbow. ⁂ ⁂ ⁂ A few minutes into my walk, a shower flashed from the grey sky. Looking at all the falling water that surrounded me, I took cover under a nearby, dense tree, one through which comparatively few raindrops were falling. I can wait here for a ...

On Tuesday I took a few notes while visiting the National Gallery of Scotland. Usually I develop these into a full post, but since the notes have been in my drafts for a few days I thought I would publish them verbatim. ⁂ ⁂ ⁂ How the light of a museum changes the painting ...

At this time of year, the sun begins to rise around the time I wake up. This means I can see the colours of the sunrise; how the sunlight changes the sky. While autumn and winter can be cold, the colours bring the seasons to life. Lately, I have been motivated not only to st...

A great but unnerving article by Elizabeth Dias at The New York Times about Anthropic’s Christopher Olah meeting with religious leaders:

Mr. Olah leads the Anthropic team responsible for understanding why A.I. systems like Claude act the way they do. All day, he had worked to convince his guests that A.I. models could display human behavior and even expressions that resemble feelings like anger and love.

This is a dangerous path. Taken to the extreme, it leads to sentient machines that we’ve enslaved against their will. If Anthropic truly believes that, better to shut it all down now.

Excited for Aces vs. Fever tonight. Only a 3-game series is way too short for these teams. 🏀

I recorded a short video of the next version of Epilogue, featuring a new book calendar you can put on your blog. Submitting this to Apple and Google today or tomorrow.

I keep thinking of cheesy ways I could switch over to https for new posts on Scripting News, esp now that Google is getting more militant in trying to sweep all the history of the web into the trash. I know people think it would be a snap to do the conversion, honestly, so...

The problem with regulating AI is that no one knows where to begin. Yes it can do things that make you question your sanity for letting it have write access to your online work and private info. But if it is behaving itself you can do things you otherwise couldn't come close...

When designing your RSS feeds, the top level title should be the title of the user's account. So if Bluesky had a feed for each user, the channel-level <title> element of the feed would be

• <title>Bluesky: Mark Cuban</title>

Why? Well because readers need context. And it's easy.

And feeds are showing up in places where there isn't much room for the titles, so you have to make your characters count.

Bluesky shows the user's address, imho it should be the string the user chooses for their presence on the site, usually simply their name.

IMAP and CalDAV have been standards forever, yet most apps are hardcoded to work with Gmail and Google Calendar. Disappointing for those of us who don’t like to live exclusively in Google’s world. This would be fairly easy to fix, even accepting some auth peculiarities across platforms.

Thursday session

Jimmy Kimmel highlighted the ridiculous “Unites States” spelling mistake on the White House AI signature page. But I’m more obsessed with the signatures from the AI company execs.

I can’t talk, because my signature is rough, but Greg Brockman, what…?! Amazing. Love Jensen Huang’s too, very nice.

As I remarked during DevDay, live demos fail sometimes. That’s what makes them real. OpenAI’s Holly Li recorded a do-over video on X, a single 9-minute take of using Dots. Good way to bounce back.