alganet

46 ms·

721.
by alganet 1y ago
I'm assuming the post is about automated "context engineering". It's not a human doing it. In this arrangement, the LLM is a component. What I meant is that it seems to me that other non-LLM AI technologies would be a be
722.
by alganet 1y ago
If I need to do all this work (gather data, organize it, prepare it, etc), there are other AI solutions I might decide to use instead of an LLM.
723.
by alganet 1y ago
My original comment included a political example, but I removed it before posting precisely to avoid confusion between what I said and simple polarization schemes. It's not about backlash. Media has several ways to deliver a different
724.
by alganet 1y ago
Dataset? That's so 2000s. Each crawl on the internet is actually a discrete chunk of a more abstractly defined, constant influx of information streams. Let's call them rivers (it's a big stream). These rivers can dry up, pres
725.
by alganet 1y ago
I don't understand your comparison.
726.
by alganet 1y ago
> But I am just pointing out that most things have a genetic pattern behind them, since all our features do. I guess I'm genetically predisposed to not give much credit to genetics then. Nothing you can do about it, I will always be
727.
by alganet 1y ago
I think you misunderstood what I said. I don't believe a 2025 version would solve the hard problem of consciousness, or even contribute meaningfully to solve it. The way I see it, the church is using _an even older_ version of the same
728.
by alganet 1y ago
Lots of things were said about AI. I could take this sort of discussion to any subject I want. The article tries to put these personalities into a "master manipulator" figure. It doesn't matter if 99% of the text actually _cr
729.
by alganet 1y ago
> a savvy negotiator and manipulator You give them too much street cred. I'm not convinced they're even good at that.
730.
by alganet 1y ago
That's not the point of the article though.
731.
by alganet 1y ago
You don't need catholics, or appeals to the beauty of sunsets, to question the "Turing Test" (there is no such thing, actually). https://courses.cs.umbc.edu/471/papers/turing.pdf Here's a quote
732.
by alganet 1y ago
This myth of Randi as some sort of icon to skeptical inquiry was largely made by the media of the time. He's a magician. He scared the shit out of some televangelists using their own tactics, and that earns him my profound respect. Hea
733.
by alganet 1y ago
> users will need new devices as artificial intelligence becomes more prevalent Some new hardware (crazy expensive, crazy features or quality or driven by semi-industrial requirements) > exploring new interaction paradigms > moving
734.
by alganet 1y ago
> "I feel a visceral feeling right now, as if someone has broken into our home and stolen something" This reminds me of something. I mean, Meta ain't no saints. However, it is funny to see an AI company complaining that so
735.
by alganet 1y ago
I don't know. It seems too vague of a description. On the guitar, I did some identifying and learning. On drums though, I just sat there and in a matter of months I was able to play to a lot of songs. I was doing polyrhythms, for examp
736.
by alganet 1y ago
This degree of vertical integration is missing from any of the AI examples you mentioned. The example you provided, the QR code generator, presents none of those qualities. You used multiple AIs from different vendors (according to you: Son
737.
by alganet 1y ago
That makes sense, thanks for clarifying it!
738.
by alganet 1y ago
Read his text more carefully: > I can't just store my prompts in git and know that I'll get the same behavior each time He's not on this idea of using english as source code. He explicitly acknowledges that it doesn't
739.
by alganet 1y ago
> uses a state machine style That's the way to go. I don't even consider other shallow and ad-hoc approaches as actually parsing it. I've been working on a state-machine based parser of my own. It's hard, I'm tar
740.
by alganet 1y ago
The kinds of genetic claims people usually make about schizophrenia are of the hereditary kind (including the post article), not random mutations. I attribute this to how the illness is researched: finding a genetic factor would be a major
741.
by alganet 1y ago
I had a guitar since I was 16, and never gave up trying to learn. I reached a plateau very quickly (knew some chords, simple songs), and could never go past it. I then spent almost 20 years in that plateau. Then, suddenly, it all started to
742.
by alganet 1y ago
Completely anectodal: Right after the time I was diagnosed (~36), I started to become weirdly good at some stuff. Music, for example. I've been playing for almost two decades and couldn't progress after a certain level. This chang
743.
by alganet 1y ago
What if someone peppers their malicious script with `# shellcheck disable=` pragmas?
744.
by alganet 1y ago
https://martinfowler.com/aboutMe.html > [...] I work for Thoughtworks [...] > [...] I don't come up with original ideas, but do a pretty good job of recognizing and packaging the ideas of others [...] > [...]
745.
by alganet 1y ago
> Yes, shell is definitely too weak to parse JSON! Parsing is a trivial, rejecting invalid input is trivial, the problem is representing the parsed content in a meaningful way. > bash completion scripts try to parse bash in bash You&#
746.
by alganet 1y ago
I see the evidence to the contrary. SpaceX took years to develop and debug their own engine, from scratch. Highly specialized work. They proud themselves of not relying on external supply chains, and doing everything in house. The refurbish
747.
by alganet 1y ago
I must add that it is also possible that many people foresaw this from the very beginning, and are working towards disrupting or minimizing potential LLM psychological effects using a variety of techniques.
748.
by alganet 1y ago
If these cases are as real as they are portrayed (big if 1), and the cause can really be attributed solely to LLMs (big if 2), then it is just a matter of time until this is weaponized. Two big ifs considered, it is reasonable to assume tha
749.
by alganet 1y ago
You need to understand that Mr. Fowler works for a consultancy. LLMs sound great for consultants. A messy hyped technology that you can charge to pretend to fix? Jackpot. All things these consultancies eventually promote are learnings they
750.
by alganet 1y ago
I was planning to do a very in-depth analysis of multiple shell interpreters, with tests, prose, tutorials, examples. Even bought the domain. Now I know AI companies would likely steal it and not give attribution. I'm not some big shot

More ›