The real killer AI app will be a box you can attach to your computer with a huge amount of memory and CPU that allows you to bake your own bot, as a bootstrap, you use the bot to build the bot you want. We're all going to architect people. You think it sounds trippy? We thought portable phones were pretty trippy in the late 80s. I kid you not. #
A young person just beginning their career wants advice on being an entrepreneur, and specifically how to judge people as worthy of trust. My answer: Assume until proven otherwise that everyone is out to advance only themself. Unless you can see actual things they've done that were for the good of others or everyone, don't believe their claims. I didn't start out this way, quite the opposite, I thought everyone wanted to do good stuff that makes the world better. We certainly need a lot more of that. But being in it for yourself is a survival trait I guess, and until we reinvent evolution to select for group-good, we're more or less permanently fucked. At the slightest sign of progress, some asshole will try to take the common good for themselves. A great example is how Big Tech has taken over the open web and is gradually turning it into a cash cow for themselves. #- There's been some recent discussion about whether ActivityPub is the standard everyone should be building on, and that new ways of social networking, or even previously existing ones, should stop and implement ActivityPub.#
- My answer is this -- if ActivityPub were sufficient, I'd say yes -- let's coalesce around one standard, and stop trying to make something slightly better. I have a motto that goes with this -- One way of doing something is better than two, no matter how much better the second way is. For a couple of reasons: 1. You still have to support both, so you just made everyone's life more complicated, assuming the second way gains traction. 2. The goal is interop, not creating standards. Both of these are part of my Rules for Standards-makers. #
- But honestly, ActivityPub is both insufficient and a mess. And I'm not even convinced that what it's doing is all that worthwhile. I don't see conversation on social nets as having much value. Replies to my posts tend to be spam, people looking for exposure, hitching a ride on a post that's getting some flow, or to be annoying. It's very rare that new ideas or information come to light this way. #
- And all the arguing in the world won't change the way things are going. ActivityPub isn't the standard. People want to be Mastodon-compatible, not ActivityPub-compatible. Much of what Mastodon is doing isn't part of ActivityPub, it's their own API. #
- I've seen it said that Atom was a good thing, that somehow RSS wasn't enough. And all the "IndieWeb" protocols that just reinvent the Metaweblog API. Depends what your goals were. If you were intending to keep RSS and the blogging APIs from growing, then they were great successes. People spent time judging not aiming for more interop, whether items had their proper names, which is a common devolution in this kind of work, they even have a name for it, bike-shedding. #
- It's better to reduce the number of competing formats than to increase the number. #
- Again all that matters is interop, and splitting an open standard into multiple competing standards works against interop.#
- Back to ActivityPub. I wish it were simpler, that you could get something interesting working quickly that delivers interop. But it isn't that way, and no one seems interested in making it so. So imho the world will not coalesce behind ActivityPub, I wish it could, but it can't and won't happen. #