AI component libraries in your site make your web app even more easily consumed and subsumed by AI chat clients.
This not only kills pages, but it kills the concept of a browser where the user agent is a human, rather than making your pages be designed where the user agent is an AI agent.
That doesn't make me happy to experience because I'm guessing that after a generation or so, web designers will not only do mobile first designs with stupid amounts of white space and not taking advantage of the desktops greater screen real estate and precise mouse movements, but AI first websites will get so popular that browsing sites manually will look like trying to use a text only browser in the JavaScript world.
Easy for me to do a depressing take, but hopefully the bitter lesson of AI will help this particular projected future not come to pass because the AI will get smart enough that it will embed a browser right there in line and just render the window for the user, or it will otherwise gets good enough at screen scraping and UI automation that it can just use an existing browser, just like a human, the sites won't be dumbed down even further for AI consumption.
at best we have walled garden content; and when those are scraped (either by the host or by more sophisticated bots) those walled gardens will hopefully rot under an inability to drive advertisement revenue.
I agree, I think we're at the edge of a paradigmatic shift away from humans navigating TCP-IP itself. What that looks like, I don't know, but given trends (like dynamic pricing, human-futures marketing, surveillance, and consolidation of computing under mega-companies) I can imagine: local beacons screaming AI advertisement components across a geospatial sneakernet. Auditorium-based ticketed podcasting and AR/VR/meatspace events. Thoughtful hackers reminiscing of better times simulating them in web-assembly driven first-person POV "sites" and a rolling set of encryption keys for read-access (just send them BTC)
without an ecosystem for humans to contribute meaningfully to a feedback loop that allows for free group assembly around like interests, monetary growth for hosts and other participants, and some degree of presence / searchability / permanence, the current text-only web page paradigm is doomed.
I am of course unqualified to provide useful commentary on it, but I find this concept to be new and interesting, so I will be watching this page carefully.
My use case is less so trying to hook this up to be some sort of business workflow ClawdBot alternative, but rather to see if this can be an eventually consistent engine that lets me update state over various documents across the time dimension.
could I use it to simulate some tabletop characters and their locations over time?
that would perhaps let me remove some bookkeeping how to see where a given NPC would be on a given day after so many days pass between game sessions. Which lets me do game world steps without having to manually do them per character.
That's a very interesting use case you brought to the table! I've also dreamt about having an agent as my co-host running the sessions. It's a great PoC idea we might look into soon.
I built out Nexus 2 years ago and have run Jfrog 4+ years ago, and now am back at the Artifact Repo choosing process again. I'm going to try to get away with AWS ECR for containerized artifacts to defer purchasing a proper versioned artifact repo, but I will be keeping a close eye on this one, I respect the work and design that went into this. I couldn't have done it better myself!
thanks for the recommendation, I've put a hold on it for my library now.
This article reminds me of another book [1] called Holacracy where how a business is run is systematized according to other pre-defined principles. David Allen, a productivity trainer, used it at his own company for several years before eventually moving away from it because the ongoing overhead to keep its system up was too much.
I wonder if this system will end up like that as well. I love the idea, but I think humans operate at a squishier level than our computers do, there's a risk of 'massive bureaucratic dehumanization and inflexible processes' and the Iron Law of Organizations that make such efforts as that book and this article fraught with peril. Taylorism has its limits.
But hey, if this works, I'll be excited to see more businesses adopting better practices and less painful fumbling around trying to do practices in an organic or unplanned way.
I'm so amazed to find out just how close we are to the start trek voice computer.
I used to use Dragon Dictation to draft my first novel, had to learn a 'language' to tell the rudimentary engine how to recognize my speech.
And then I discovered [1] and have been using it for some basic speech recognition, amazed at what a local model can do.
But it can't transcribe any text until I finish recording a file, and then it starts work, so very slow batches in terms of feedback latency cycles.
And now you've posted this cool solution which streams audio chunks to a model in infinite small pieces, amazing, just amazing.
Now if only I can figure out how to contribute to Handy or similar to do that Speech To Text in a streaming mode, STT locally will be a solved problem for me.
Happy to answer questions about this (or work with people on further optimizing the open source inference code here). NVIDIA has more inference tooling coming, but it's also fun to hack on the PyTorch/etc stuff they've released so far.
And here I was wondering if someone was plotting their tabletop game scenario plots or creative writing plots using claude to do some bookkeeping sonI was the creatorncould focus on brainstorming in a writers room of 1+1.
Oh well, still an interestign article that shows statistics can be posed in such a way to say anything about anything if you squint them.
Agreed, (heirarchical if you must) state machines consuming queues and writing to queues via messages wins. If you are FP minded like me, then you are set up to cleanly separate IO to the edges and have a functional core imperative shell hexagonal architecture for less additional overhead thsn a standard java beans style OO logical design.
School choice works within municipalities. A huge part of the point of modern inner-ring suburbs is to allow wealthier people to exempt themselves from that system.
This not only kills pages, but it kills the concept of a browser where the user agent is a human, rather than making your pages be designed where the user agent is an AI agent.
That doesn't make me happy to experience because I'm guessing that after a generation or so, web designers will not only do mobile first designs with stupid amounts of white space and not taking advantage of the desktops greater screen real estate and precise mouse movements, but AI first websites will get so popular that browsing sites manually will look like trying to use a text only browser in the JavaScript world.
Easy for me to do a depressing take, but hopefully the bitter lesson of AI will help this particular projected future not come to pass because the AI will get smart enough that it will embed a browser right there in line and just render the window for the user, or it will otherwise gets good enough at screen scraping and UI automation that it can just use an existing browser, just like a human, the sites won't be dumbed down even further for AI consumption.