Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The meeting of the minds that launched AI

spectrum.ieee.org

Post reply on HN
Print view View original

41–50 of 59 posts

Re: The meeting of the minds that launched AI

#41

Post by gumby »

dontreact wrote: post #33

This group of people may have been the first to mention the words “AI” prominently in academia but is this flag planting or are they truly foundational to the work with the same name today? If none of these people had done anything, would we really be far behind? My sense is that modern AI has more to owe Fukushima’s neocognitron, Hubel and Wiesel, and the connectionists than any intellectual descendant of the work m…

Are you trolling? Minsky, McCarthy, Newell and Simon all went on to win Turing awards for their work, (as later did several other AI luminaries over the decades). And Claude Shannon?

In the mid sixties Minsky and Papert published a paper/book called "Perceptrons" which explored the limit of perceptrons, though it said those limits could be overcome by multilayer networks, if they were ever computationally feasible. And, thanks to Moore's law, they now are.

Almost every one of those attendees is a well known scholar whose work you depend on every day outside AI (as well as within it, though nobody really depends NNs yet). Of them, the least known in computing, Ray Solomonoff, might be the smartest of the bunch, though he was never univerity affiliated as far as I know.

Re: The meeting of the minds that launched AI

#42

Post by belugacat »

famouswaffles wrote: post #22
fnordpiglet wrote: post #16

Earlier quoted context omitted.

LLMs of today display amazing abductive abilities but are limited in inductive and deductive abilities, as well as other optimization techniques of classical AI and algorithms. These abductive abilities are unique and exciting because we’ve typically done really poorly with ambiguous and complex semantic spaces like this. However I think the excitement has obscured the fact it’s just a piece of a larger machine. Why…

>but are limited in inductive and deductive abilities LLMs are great at induction. In a broad sense, they are also very good at deduction. "I define a new word, the podition. A podition is any object that can fit on a podium. Is a computer a podition ? Why ?" A correct answer is deductive. LLMs eat these kind of questions for breakfast. Even the OG 2020 GPT-3 could manage them. You really do have to stretch deduction…

For breakfast, you say?

User:

I define a new word, the podition. A podition is any object that can fit on a podium. Is a computer a podition ? Why ?

ChatGPT:

The term "podition" is a newly coined word, and its definition as "any object that can fit on a podium" is quite specific. Whether or not a computer can be considered a "podition" depends on the size and design of the podium in question, as well as the specific characteristics of the computer.

If the podium is large enough and designed to accommodate objects of various sizes, then a computer could potentially be considered a podition if it fits on the podium without exceeding its space or causing any issues. However, if the podium is small or not designed to hold objects like a computer, then it may not qualify as a podition.

In summary, whether a computer is a podition or not depends on the podium's size and design and whether the computer can comfortably fit on it without causing any problems. The term "podition" itself is a new and specific concept, so its application may vary depending on context.

User:

Using common sense definitions of a computer and a podium, is a computer a podition?

ChatGPT:

Using common sense definitions of a computer and a podium, a computer is typically not considered a "podition." A computer is an electronic device used for processing and storing data, while a podium is a raised platform or stand typically used for holding notes, books, or supporting a speaker during a presentation. These two objects serve very different purposes and have distinct characteristics.

A computer is not designed to be placed on a podium, and it would generally be impractical to do so. A podium is meant for supporting objects like papers, books, or notes during a presentation, not electronic devices like computers. Therefore, in common usage and understanding, a computer would not be classified as a "podition."

Re: The meeting of the minds that launched AI

#43

Post by famouswaffles »

belugacat wrote: post #42
famouswaffles wrote: post #22

Earlier quoted context omitted.

>but are limited in inductive and deductive abilities LLMs are great at induction. In a broad sense, they are also very good at deduction. "I define a new word, the podition. A podition is any object that can fit on a podium. Is a computer a podition ? Why ?" A correct answer is deductive. LLMs eat these kind of questions for breakfast. Even the OG 2020 GPT-3 could manage them. You really do have to stretch deduction…

For breakfast, you say? User: I define a new word, the podition. A podition is any object that can fit on a podium. Is a computer a podition ? Why ? ChatGPT: The term "podition" is a newly coined word, and its definition as "any object that can fit on a podium" is quite specific. Whether or not a computer can be considered a "podition" depends on the size and design of the podium in question, as well as the specific…

>For breakfast, you say?

Yes and it's honestly pretty funny that you think this proves otherwise. The error is mine. a computer is not always small enough to fit on a podium. It's not about a "common sense" definition or not. and GPT caught that.

Go ahead and replace computer with any object that is always small enough to fit on a podium.

Re: The meeting of the minds that launched AI

#44

Post by belugacat »

famouswaffles wrote: post #43
belugacat wrote: post #42

Earlier quoted context omitted.

For breakfast, you say? User: I define a new word, the podition. A podition is any object that can fit on a podium. Is a computer a podition ? Why ? ChatGPT: The term "podition" is a newly coined word, and its definition as "any object that can fit on a podium" is quite specific. Whether or not a computer can be considered a "podition" depends on the size and design of the podium in question, as well as the specific…

>For breakfast, you say? Yes and it's honestly pretty funny that you think this proves otherwise. The error is mine. a computer is not always small enough to fit on a podium. It's not about a "common sense" definition or not. and GPT caught that. Go ahead and replace computer with any object that is always small enough to fit on a podium.

The mental gymnastics are beautiful and entertaining.

Google image "podium". Google image "computer". The vast majority of computers shown can be placed on the vast majority of podiums shown.

But again, as the other commenter points out, this is abductive reasoning.

In addition to getting lost in the irrelevant weeds of "these two objects serve very different purposes and have distinct characteristics", ChatGPT also fails to point out that there are 2 common definitions for podium - one made for lecturing, on which you could place most modern computers that aren't 30kg hunks of metal (all of my university lecturers definitely had computers on the podium). And one made for humans, on which you could definitely place any bulky desktop computer short of a mainframe.

What this really shows though is that LLM output is a mirror in which people will read mostly what they want to read which matches their narrative.

Re: The meeting of the minds that launched AI

#45

Post by ggm »

As a 7 year old, I met two of them at IFIP68 in Edinburgh. I don't have much recall of Minsky, but McCarthy was nice. We went out to turnhouse airport and he flew a light plane around. At dinner he commented we did the washing up by hand and when he offered to send us a dishwasher my mother (of course) said "no no John don't be silly" but she said to me in the kitchen later on: "if he ever offers again accept"

We also ate at a chinese restaurant across town and he was surprised when me and my brothers and sisters said birds-nest soup was disgusting.

Susy McCarthy stayed with us for a while. She was a typical Californian teenager, ate her cereal with the wrong hand holding the spoon and her feet up on the chairs and we younger kids were told in no uncertain terms not to emulate her. She refused to go to school and there were exasperating phone calls about what to do. In the end she went somewhere else. Johns wife died climbing Mt Everest. it was really sad. I think it was the first all-women's climbing team to try the route (or even Everest's pinnacle)

Minsky and McCarthy were quite happy to respond to emails in the 80s and 90s. I asked Minsky about writing with Harry Harrison and he said he was frustrated at the ending of the book they did ("the Turing Option") and McCarthy talked about his ideas of long term information ontologies in a notation or language he called "Elephant"

McCarthy was a notable USENET curmudgeon and would be oppositional to any anti-nuclear, anti-oil, anti-plastic agenda he saw. He also said he cried when he learned the bombs had been dropped, and that the invasion of Japan was off, because he knew he wouldn't die on a beach in an opposed landing. (It is possible I have this mixed up with memories of what Paul Fussel wrote, but I am pretty sure John said this)

John played "Alice's Restaurant" with a guitar, he had a very nasally whiny voice.

Re: The meeting of the minds that launched AI

#47

Post by famouswaffles »

belugacat wrote: post #44
famouswaffles wrote: post #43

Earlier quoted context omitted.

>For breakfast, you say? Yes and it's honestly pretty funny that you think this proves otherwise. The error is mine. a computer is not always small enough to fit on a podium. It's not about a "common sense" definition or not. and GPT caught that. Go ahead and replace computer with any object that is always small enough to fit on a podium.

The mental gymnastics are beautiful and entertaining. Google image "podium". Google image "computer". The vast majority of computers shown can be placed on the vast majority of podiums shown. But again, as the other commenter points out, this is abductive reasoning. In addition to getting lost in the irrelevant weeds of "these two objects serve very different purposes and have distinct characteristics" , ChatGPT also…

>The mental gymnastics are beautiful and entertaining.

Sure

>But again, as the other commenter points out, this is abductive reasoning.

Deduction is divining the effect from rule and cause. The rule is the definition of the podition, the cause is that a computer is small enough to fit a podium and the effect is that a computer is a podition.

>Google image "podium". Google image "computer". The vast majority of computers shown can be placed on the vast majority of podiums shown.

Deduction does not deal with probable statements. The rule and cause must both be true. My cause is not always true.

Like i said, i genuinely do think it's funny that you read all that and decided it was absence of reasoning. Just kind of shows the state of things.

Re: The meeting of the minds that launched AI

#48

Post by fnordpiglet »

famouswaffles wrote: post #40
fnordpiglet wrote: post #37

Earlier quoted context omitted.

This is actually abduction. Induction doesn’t fail, abduction is probabilistic.

three components: a rule, a cause, and an effect. The example i gave is deductive. i.e divining the effect from rule and cause. The rule is the definition of the podition, the cause is that a computer is small enough to fit a podium and the effect is that a computer is a podition. Induction is divining the rule from the cause and effect. It can definitely fail. There could be exceptions to the general rule that aren'…

I’m sorry you’re right, I dashed off a quick reply and wrote the wrong word as I was distracted.

However I still stand by the statement that their inductive and deductive reasoning is weaker than abductive. This is why they so easily hallucinate - that the very nature of choosing the semantically likely next token is at its root abductive.

GPT is remarkable, but it’s not reasoning in any meaningful sense. It’s not structuring logical constructs and drawing conclusions. I’d hold by my assertion that they are abductively simulating inductive and deductive reasoning.

Re: The meeting of the minds that launched AI

#49

Post by blast »

ggm wrote: post #45

As a 7 year old, I met two of them at IFIP68 in Edinburgh. I don't have much recall of Minsky, but McCarthy was nice. We went out to turnhouse airport and he flew a light plane around. At dinner he commented we did the washing up by hand and when he offered to send us a dishwasher my mother (of course) said "no no John don't be silly" but she said to me in the kitchen later on: "if he ever offers again accept" We als…

> As a 7 year old, I met two of them at IFIP68 in Edinburgh

Can you fill in some background? What was IFIP68 and how did you end up there as a 7 year old?

Post reply on HN