B.T. Franklin
@btfranklin.info
software architect // applied AI technologist // computational creativity researcher // futurist // world traveler
https://btfranklin.info
https://btfranklin.info
Sure, but I'm saying even that is inconsistent. I will often just say "you have a skill related to this, so you might engage it" if I know there's one I'd like for it to consult. I am quite confident this will not be a problem for long.
January 21, 2026 at 1:16 AM
Sure, but I'm saying even that is inconsistent. I will often just say "you have a skill related to this, so you might engage it" if I know there's one I'd like for it to consult. I am quite confident this will not be a problem for long.
Yeah, I kind of figured that out by observing Codex's behavior. It sometimes "chooses" to read the AGENTS.md, but most of the time it doesn't, unless I ask it to. I kind of find the same thing with skills. It also sometimes decides to read AGENTS.md kind of mid-stream, as if confirming approach.
January 20, 2026 at 11:57 PM
I've been coding for my entire adult life, and this doesn't "disturb" me at all. The actual writing of the code is the most annoying part. I want to *build* things that do cool shit. Typing the syntax can give you a feeling like winning at Tetris, but it's not the Big Win.
January 20, 2026 at 3:43 AM
I've been coding for my entire adult life, and this doesn't "disturb" me at all. The actual writing of the code is the most annoying part. I want to *build* things that do cool shit. Typing the syntax can give you a feeling like winning at Tetris, but it's not the Big Win.
What you'll get from that is a moderately-realistic, imperfect foundation from which you can build "real" software. None of this, to me, is about what I can "one shot". 5/5
January 20, 2026 at 12:38 AM
What you'll get from that is a moderately-realistic, imperfect foundation from which you can build "real" software. None of this, to me, is about what I can "one shot". 5/5
The next thing I do with a project at that point is tell the model "Here is a prototype. We want to productize it, so it's going to need to be refactored to have an architecture that incorporates best practices, is maintainable, future-proof, and sane." 4/x
January 20, 2026 at 12:36 AM
The next thing I do with a project at that point is tell the model "Here is a prototype. We want to productize it, so it's going to need to be refactored to have an architecture that incorporates best practices, is maintainable, future-proof, and sane." 4/x
That gets you to an initial prototype, which is architecturally a mess but also simple. A few refinements to "see if it can work", and then you can decide whether to go further with it. If it's not working, you scrap it. If it holds promise, then you don't have to start from zero! 3/x
January 20, 2026 at 12:34 AM
That gets you to an initial prototype, which is architecturally a mess but also simple. A few refinements to "see if it can work", and then you can decide whether to go further with it. If it's not working, you scrap it. If it holds promise, then you don't have to start from zero! 3/x
For example, I have found great success by asking for an initial prototype of a project idea fairly loosely-specified, but representing the surface-level vision I'm after, and built on the technology I'd want to use if it were "real". 2/x
January 20, 2026 at 12:33 AM
For example, I have found great success by asking for an initial prototype of a project idea fairly loosely-specified, but representing the surface-level vision I'm after, and built on the technology I'd want to use if it were "real". 2/x
I've been doing this, too. It's such a miraculous experience, and it changes the way I evaluate ideas before I ever touch a keyboard. I can let the smaller ideas in, because I know they're not going to be pushing a "better idea" out of my queue.
January 20, 2026 at 12:15 AM
I've been doing this, too. It's such a miraculous experience, and it changes the way I evaluate ideas before I ever touch a keyboard. I can let the smaller ideas in, because I know they're not going to be pushing a "better idea" out of my queue.
One thing about language is, you can't invent a word and then "own it" and redefine it unilaterally for eternity. It does not mean what he used it for here.
January 18, 2026 at 9:58 PM
One thing about language is, you can't invent a word and then "own it" and redefine it unilaterally for eternity. It does not mean what he used it for here.
Reposted by B.T. Franklin
"enshittfying our potable water" truly is the enshittification of "enshittification"
January 18, 2026 at 6:42 PM
"enshittfying our potable water" truly is the enshittification of "enshittification"
And yet, annoyingly, it's asserted as if a foregone conclusion. There's a lot of performative "membership language" throughout.
January 18, 2026 at 9:53 PM
And yet, annoyingly, it's asserted as if a foregone conclusion. There's a lot of performative "membership language" throughout.
Sounds like maybe what she meant was "I want to watch the bears game."
January 18, 2026 at 9:07 PM
Sounds like maybe what she meant was "I want to watch the bears game."
Making assertions like "humans possess intelligence and machines do not" without *any* claimed basis for that statement instantly removes credibility. I mean, I can just respond to it like this: "Yes, machines do possess intelligence." Who wins?
That's also the wrong use of "enshittifying."
That's also the wrong use of "enshittifying."
January 18, 2026 at 8:58 PM
Making assertions like "humans possess intelligence and machines do not" without *any* claimed basis for that statement instantly removes credibility. I mean, I can just respond to it like this: "Yes, machines do possess intelligence." Who wins?
That's also the wrong use of "enshittifying."
That's also the wrong use of "enshittifying."
Does this mean that the goofy "tainted by AI" list will have to add Linux to it? LOL Good luck with that one, folks!
January 11, 2026 at 8:06 AM
Does this mean that the goofy "tainted by AI" list will have to add Linux to it? LOL Good luck with that one, folks!
This is the "big point" that I haven't seen very many people making. Humans have all had the ability to do lots of things for themselves for a long time, but we still would rather pay to have those things done by others who care more to do so.
January 10, 2026 at 6:02 PM
This is the "big point" that I haven't seen very many people making. Humans have all had the ability to do lots of things for themselves for a long time, but we still would rather pay to have those things done by others who care more to do so.
I have mine set to go to Perplexity, so I use it quite a lot. But the various "portal-style" classic homepages are of no use to me.
January 7, 2026 at 7:30 PM
I have mine set to go to Perplexity, so I use it quite a lot. But the various "portal-style" classic homepages are of no use to me.
I mean, when you really boil it down, this is just what fear looks like. Fear is pre-rational. I don't like it any more than you do, but also I guess we'd better try to see it for what it is.
December 28, 2025 at 5:52 PM
I mean, when you really boil it down, this is just what fear looks like. Fear is pre-rational. I don't like it any more than you do, but also I guess we'd better try to see it for what it is.
He was able to identify the giraffe. Allegedly.
December 27, 2025 at 1:04 AM
He was able to identify the giraffe. Allegedly.
The whole "wall" narrative feels more and more to me like a desperate prayer from people who really *need* it to be true in order to feel okay about the world.
December 16, 2025 at 1:21 AM
The whole "wall" narrative feels more and more to me like a desperate prayer from people who really *need* it to be true in order to feel okay about the world.
An illiterate busting out the emdash...just saying.
December 2, 2025 at 12:52 AM
An illiterate busting out the emdash...just saying.
Mark Kelly for President 2028
December 2, 2025 at 12:51 AM
Mark Kelly for President 2028