A.I. (1981)
newyorker.com
newyorker.com
> But someday, according to Minsky, the most useful programs for personal computers will be based on artificial-intelligence programs that write programs of their own. The idea is for an ordinary person — not a programmer — to describe what he wants a program to do in informal terms, perhaps simply by showing the program-writing program some examples.
> Then it will write a computer program to do what was described — a process that will be much cheaper than hiring a professional programmer.
They still need a programmer to give proper instructions, though. I don't think an end user can prompt them to write anything remotely sophisticated the current technology.
When these tools are able to detect ambiguity in specifications and start asking the right questions to clarify those ambiguities, they will also start being useful for non-programmers.
that has pretty broad applications for all sorts of non-programmers.
yeah, you're right if the goalpost is "please make a HIPAA compliant DBS from scratch for use in California", but most non-programmers aren't really interested in specification compliant programs that meet meticulous measures of rigor; they just want things that simplify their routine when cashing out a customer, organizing their pirated movies, perusing porn sites, whatever it is that people do.
Programming as a metaphor for telling a machine what to do was a great gatekeeper for people for many years, chatgpt and the like are reducing the 'fee to entry' immensely for non-technical people, and I say that as someone who is witnessing non-technical people in my life accomplish things they would have never tried previously.
How much of this is thanks to hundreds of tutorials / question & answers on stack overflow about these tasks?
As AI supplant the sites that host these QnA and professional tutorial writers - even amateur writers might feel miffed that their efforts will be regurgitated by AI with no credit given - how will AI keep up the quality of their responses?
If humans can train themselves without AI, then so can AI train without humans.
I’ve used it to do this very thing and while it was helpful, it really struggled with sticking to either v1 or v2 of the language. It felt like playing whack-a-mole. I would tell it some line was incorrect syntax for v2 and it would apologize, fix that line and break something else.
The bones of the script were good, but I had to do quite a bit of editing.
I don’t feel all that negative about it though because I understand that it will only get better.
I'm told GPT-4 is much better at bug fixes, and there are various options for longer context windows.
Knowing what question to ask, though? The AI seems to contain the knowledge to all the questions, as if I ask them for a list of questions I get a useful response; and now someone has to answer the meta-question of "how do you want to present those questions to a user?" as there are many ways…
Why spend time researching and vetting some dependency hell when GPT made 90% of what I needed in a few minutes.
GPT for sure changes the build-vs-buy equation
We can think of AI as the quest for a computer program that is a complex adaptive system that can exchange information with us in an anthropomorphised manner.
LLMs are much closer to such a technology than a program that can optimally play a board game by evaluating each position and pruning its model accordingly.
No, thank you. Let us just stick to "a system that can solve problems potentially replacing a professional", which also would imply "providing a professional output".
That «anthropomorphised» could make one think of the joke (which exploits the gender-bent use of 'man' in demotic):
Seller: "This computer reasons just like a man". Female customer: "Can I have one that just reasons?"
Similarly, it should "exchange information well" - not "like a human".
"Professional" and "well" are small targets for tech firms attempting to explain R&D to their share holders.
The pursuit of artificial intelligence is something with far deeper roots and wider ranging consequences than most pedants can imagine.
Oh, the same as today, had it also happened around that time that licence got widespreadly felt as a prevailing right.
> share holders
If you want to sell, surely there is market for all kind of junk. We also have this need for the "real things" though.
> pedants
No, not pedantry: the difference is radical, between form and effectiveness. In AGI, it's the proper reasoning that makes the "real" difference - use of natural language could be useful. We want the consultant to be competent and smart, not dressed attractively and "smart".
Thank you for persisting.
The ability of AI to mass produce art reduces its value, turning what might seem to have artistic value into first design and ornamentation and finally just background noise.
A lot of anti-AI rhetoric, both the rhetoric saying that it's not possible, or that whatever new AI thing "isn't real AI", or the rhetoric that says AI is dangerous, I'm fairly sure is rooted in a probably vague sense in most people that _what they are_ is not a collection of atoms, that their conscious experience of the world is not some emergent property of neurons, but that they are, deep down, a spiritual, immaterial being, which also holds up the possibility of being _eternal_. As AI sort of chips away at the frontiers of intelligence, and shows that deep down, our experience of the world, our consciousness, all of it is an ugly bag of party tricks, it provokes a profound sense of _unease_. You're not just sort of threatening to dethrone humanity's place in the center of the universe, it's also in a sense stealing _eternal life_ from people. That's fairly heavy.
It's also why people who work on AI or work with AI tend to also be intensely interested in life extension.
(It's unfortunately clipped on mobile!)
I would have laughed at how technically uninformed this opinion seems to be. Funny how vague macroscopic reasoning fared better than a detailed bottom-up reasoning with quantification of the challenge.
OTOH I'm sure the New Yorker also wrote a lot of interesting to think about and not so well founded articles that didn't fare well.
This seems to be the pattern for the human champions who are the first to lose to computers. What a truly singular moment in existence to experience. Both a personal loss and a loss for the whole human species, and you have to bear it alone.
Anyone guess what he would think about the current generation of popular tools?