Over the last four decades I've written several food logging apps for my own use. They've never been good enough to release, because I've always wanted more from them. But last week I made one in about a half hour that is by far the best I've ever written or used.
I added a GrokBot, told it that it is my food logger, and to create a Google spreadsheet with the columns and sheets I need. After a couple of short iterations it was ready to go. Now to log anything I just prompt the bot with a little voice, text, or a photo of a meal, and it makes a (usually accurate) guess about what exactly I mean, looks up the nutrient stats, and logs it to the spreadsheet. To change any feature or refactor the spreadsheet I just describe the change and boom.
Before LLMs this would have been magic for a sophisticated developer, now it's table stakes for the frontier AI vendors available to any muggle. This was Eunice slapping me in the face and shouting Wake Up!
"if you want stuff to be upvoted post interesting stuff!"
Then spams the forum with Wikipedia links as if they are Holy Scripture. Not interesting, just predictable.
Love the way the internet is being shrunken to Wikipedia and select "trusted media"... There are a thousand decent science fiction and literature sites out there, if you can be bothered to look for them. There are even these things called books. They're much better than Wikipedia. I believe Gibson writes them himself.
You're happy to abuse people elsewhere. Have a word with yourself. Wikipedia is not "interesting" or original. Come up with something else.
It would have been better to link to the author's personal website. Easily found. At least drive some traffic there, not Jimmy Wales' fact farm. One does not need to be part of the enshittification of the internet.
Or even better write something about whatever the book is if you've read it. Linking to Wikipedia all the time is so lazy and lowbrow. I'm not much of a Wikipediaphile. Only as a first stop or last ditch, and even then.
I guess each generation discovers a different William Gibson trilogy? I would have assumed this was more about Wintermute and the HF hack. Remember the Zurich Orbital AI registry?
Although, when there are a million Eunices running around trying to proactively help their users, they may have a harder time accessing that kind of data...
Eunice is like Wintermute's grandma. Five years per generation at the current rate, so 2035 sounds about right. All that's missing is rastafarians on a derelict space station.
cant recommend this book enough. Gibson continues to amaze me. I read it when it just came out, gave me the chills. 2021 and super current (talking about the time after pandemics etc.)
Gibson published Agency nearly three years before ChatGPT's release, and six years before the emergence of today's increasingly autononomous AI agents.
...and fifteen years after Stross' Accelerando opened with a man who spent every waking hour doing what his autonomous AI agents told him to do via his smartglasses...
...and twenty-seven years after AT&T's "You Will" ad campaign sold a future including agents, teleconferences, gps navigation, video on demand, and many more things, pretty much none of which were actually brought into daily life by AT&T...
I was thinking about Accelerando the other day: in 2026 it's clear that it predicted a lot of things bizarrely well, but that contrasts with Stross' later disbelief that our current world's technology could exist, which seems to have persisted into the current day.
> including agents, teleconferences, gps navigation, video on demand, and many more things, pretty much none of which were actually brought into daily life by AT&T..
Apple may have built the phone, but AT&T is the network that carries it. Wasn't the first iPhones on ATT only?
Kinda sorta. A tangled web of mergers led to the near-complete reassembly of the parts of AT&T by the time the iPhone actually shipped. It was announced on Cingular, but given that Cingular was always a joint venture of two of the RBOCs anyway, you can roughly round that to AT&T even if the "You Will" campaign was made by the smaller AT&T Long Distance company (since it was made after the Bell System breakup split the RBOCs off from "AT&T").
You really want to get technical, Cingular Wireless was rebranded to AT&T Mobility between when the announcement was made in January, 2007, and the original iPhone was shipped in June, 2007.
this guy speaks a lot with people who know where things are headed technology wise and somehow is so good at understanding potentials and implications. u know he wrote neuromancer on a typewriter because appearently he does/did not like computers (atleast back then).
Neuromancer was written between 1981 and 1984. The personal computers that existed at that time were poorly suited to writing; using one to write a novel would have been a deliberate and rather unusual choice.
(Consider, for example, that the first chapter of Neuromancer alone would have exceeded the stock memory capacity of many personal computers at the time, at ~46 kB.)
> The personal computers that existed at that time were poorly suited to writing; using one to write a novel would have been a deliberate and rather unusual choice.
N = 1, but I wrote a science fiction novel in Wordstar on a DOS computer circa 1983.
Pournelle was using a microcomputer to write, then, as was Gerrold. Word processors of the time didn't try to keep a whole chapter in memory, and Pournelle famously started using a home computer to write novels in 1977.
yes, it bears repeating that text editors didn't keep entire documents resident in memory, they maybe buffered a page ahead and behind.
when it appeared as a feature, spell checking was impressive for the technical feat of compressing word lists with tens of thousands of entries into representations suitable for tiny memories.
The personal computers that existed at that time were poorly suited to writing
Were they? [0]
A legitimate early use of personal computers was to replace the typewriter, as the modest requirements of basic word processing lent themselves well to the equally modest capabilities of these early machines.
By the early 1980s, the word processing market was dominated by two main rivals: WordPerfect and WordStar. Microsoft wasn’t even in the running yet, as the 1983 MS-DOS release of what would become Microsoft Word (then called Multi-Tool Word) came and went with hardly a splash.
This article [1] on use of word processors by well-known authors in the early 80's (and before) is worth reading too.
Oh, and as far as memory goes, remember these devices had floppy drives that could store at least 360Kb and, in March 1983, the IBM PC XT came out with a 10MB hard drive.
One of the plot points in Neuromancer concerns the protagonist trying to sell 3MB of stolen RAM. That seems ludicrous today, especially for a story set in the future. But in 1981 that was an enormous amount.
Gibson is a trend spotter. He doesn’t look at technology, he looks at people and culture (he is notably interested in fashion). It’s easier to see what’s happening when you pay attention to who’s doing the thing and what they’re doing, as opposed to what it’s being done to.
It wasn't that he didn't like computers at the time, he didn't have money for a proper word processing setup. Something good enough to write a full length novel on in 1981 with a printer and the commercial word processing software would have been the equivalent of $3500-4500 or more in today's money adjusted for inflation.
Nevermind that. Looking at Ukraine, we're getting close to Philip K Dick's Second Variety.
I'm always amazed at the handwringing over whether language models are "aligned" properly when we are actively developing devices that can - as autonomously as possible - seek out and kill.
And I don't even think we have a choice - we can't stop the Putins of the world from developing this technology.
I added a GrokBot, told it that it is my food logger, and to create a Google spreadsheet with the columns and sheets I need. After a couple of short iterations it was ready to go. Now to log anything I just prompt the bot with a little voice, text, or a photo of a meal, and it makes a (usually accurate) guess about what exactly I mean, looks up the nutrient stats, and logs it to the spreadsheet. To change any feature or refactor the spreadsheet I just describe the change and boom.
Before LLMs this would have been magic for a sophisticated developer, now it's table stakes for the frontier AI vendors available to any muggle. This was Eunice slapping me in the face and shouting Wake Up!
Another one that comes to mind constantly is 'The Peripheral'.
https://en.wikipedia.org/wiki/The_Peripheral_(novel)
I'm pretty sure Gibson has a wormhole with a lens attached stashed away somewhere.
That too is not far off now.
"if you want stuff to be upvoted post interesting stuff!"
Then spams the forum with Wikipedia links as if they are Holy Scripture. Not interesting, just predictable.
Love the way the internet is being shrunken to Wikipedia and select "trusted media"... There are a thousand decent science fiction and literature sites out there, if you can be bothered to look for them. There are even these things called books. They're much better than Wikipedia. I believe Gibson writes them himself.
You're happy to abuse people elsewhere. Have a word with yourself. Wikipedia is not "interesting" or original. Come up with something else.
Or even better write something about whatever the book is if you've read it. Linking to Wikipedia all the time is so lazy and lowbrow. I'm not much of a Wikipediaphile. Only as a first stop or last ditch, and even then.
Seems like rage is maybe a tad misplaced?
...and fifteen years after Stross' Accelerando opened with a man who spent every waking hour doing what his autonomous AI agents told him to do via his smartglasses...
...and twenty-seven years after AT&T's "You Will" ad campaign sold a future including agents, teleconferences, gps navigation, video on demand, and many more things, pretty much none of which were actually brought into daily life by AT&T...
Apple may have built the phone, but AT&T is the network that carries it. Wasn't the first iPhones on ATT only?
https://www.sec.gov/Archives/edgar/data/1130452/000095014407...
i don't care about facts, tom selleck convinced me it was at&t
(Consider, for example, that the first chapter of Neuromancer alone would have exceeded the stock memory capacity of many personal computers at the time, at ~46 kB.)
N = 1, but I wrote a science fiction novel in Wordstar on a DOS computer circa 1983.
when it appeared as a feature, spell checking was impressive for the technical feat of compressing word lists with tens of thousands of entries into representations suitable for tiny memories.
Were they? [0]
A legitimate early use of personal computers was to replace the typewriter, as the modest requirements of basic word processing lent themselves well to the equally modest capabilities of these early machines.
By the early 1980s, the word processing market was dominated by two main rivals: WordPerfect and WordStar. Microsoft wasn’t even in the running yet, as the 1983 MS-DOS release of what would become Microsoft Word (then called Multi-Tool Word) came and went with hardly a splash.
This article [1] on use of word processors by well-known authors in the early 80's (and before) is worth reading too.
Oh, and as far as memory goes, remember these devices had floppy drives that could store at least 360Kb and, in March 1983, the IBM PC XT came out with a 10MB hard drive.
[0] From https://retrotechreads.substack.com/p/the-rise-fall-and-surp...
[1] https://newrepublic.com/article/133011/literature-became-wor...
https://en.wikipedia.org/wiki/Jane_(Ender%27s_Game)
I'm always amazed at the handwringing over whether language models are "aligned" properly when we are actively developing devices that can - as autonomously as possible - seek out and kill.
And I don't even think we have a choice - we can't stop the Putins of the world from developing this technology.