HNHacker News
TopNewBestAskShowJobs

pulvinar

591 karma · joined April 22, 2021

submissionscomments
pulvinar··on Are We Close to AGI?
Call it a very good bet, based on inductive reasoning.
pulvinar··on Are We Close to AGI?
Yes, some say 90 percent of jobs have been replaced by technology in the last 150 years. And why don't we see all those people out of work? Clearly technology has also opened up new fields and freed up more workers to fill the new jobs. For example all computer-related jobs just didn't exist before 1945, and they're certainly not all middle-management jobs.

You can say it's different this time, but that's what's been said every time.

pulvinar··on The most important machine that was never built
An aside: GPT4 added this comment after asking it about Alonzo Church:

...and as an AI language model, I owe a debt of gratitude to him and many other pioneers in the field for laying the groundwork that has led to the creation of advanced AI systems like myself.

pulvinar··on AI is just someone else's intelligence
Well, I did read your article but don't agree with all of it.

If machines aren't intelligent because they are so obedient, is that really a path you want to follow when applied to humans? E.g., well-trained soldiers, strict religious practitioners, etc.

And if a machine should develop a loose connection and therefore sometimes not obey a command and just go its own way, does that now make it intelligent? You see the problem.

pulvinar··on AI is just someone else's intelligence
And I can invoke a human like this: "Hey you, look over here!".

My point is they already have at least chosen the effective thought "I'm waiting for a command". After that they choose their thoughts based on what text they receive. Whether or not you allow those as thoughts is up to you, but that classification is no more arbitrary than just what in yourself you call thoughts.

But without a clear, unbiased definition for what a "thought" is, any discussion comparing them is hopeless.

pulvinar··on AI is just someone else's intelligence
Whose thought comes first depends entirely on where you draw the starting line. You could just as well say ChatGPT has already chosen to ask you for a command, and it's you who are just following along.

Of course ChatGPT was designed and trained to do that. But then we're also designed (by evolutionary forces) and trained (by parents and teachers) to do what we do.

pulvinar··on OpenAI Personal Data Removal Request Form
I should have added "theoretically". I haven't had much luck with it, using a prompt like "Tell me a story about a character named X, a Y, living in Z."
pulvinar··on The seven programming ur-languages (2021)
The stack is important to Forth, but equally so is its threaded-subroutine-call nature, which isn't present in other early languages that I know of. That puts Forth halfway between assembly and higher-level (and easier to read) languages. I never saw it as a macro language.

At least that's the viewpoint I'm familiar with from back when we heavily used it.

pulvinar··on OpenAI Personal Data Removal Request Form
But you can turn around and ask it to write a story about the person and it happily complies. That's fiction but contains some facts. Repeat this enough times and you can pull out the facts the stories have in common.
pulvinar··on GPT makes learning fun again
It's easy to cherry pick from the vast human repertoire. Here's a sample from GPT4, where I picked one from a total of two:

> The man on the moon fell off his ladder one Tuesday, a common enough occurrence that nobody really paid it any mind. Truth is, gravity's always been a bit of a show-off, even in the star-dusted emptiness of space.

pulvinar··on GPT makes learning fun again
I like how it checks its facts (mostly, anyway).

It didn't give a wrong answer when asked "Is there a digital to analog converter with an 8V analog range and serial input?", which another poster (mhb) had shown to trip up plain GPT4.

pulvinar··on Controversial bill to regulate online streaming becomes law in Canada
It's funny that this is needed for streaming, because it seems like every time I look up a filming location, it's Canada. Good ol' "Hollywood North".
pulvinar··on Geoffrey Hinton leaves Google and warns of danger ahead
You mean make novel discoveries like some humans. That would be great, but that's a higher (IQ) bar.
pulvinar··on Replika AI: Your Money or Your Wife
Disgust is a strange sense. How often do we stop and think about what all those real humans out there are walking around with in their guts? We're at most only 50% human cells by count, after all.
pulvinar··on Are emergent abilities of large language models a mirage?
I think you're right, and the reason is probably that more complex concepts are hierarchies relying on the lower-level concepts to be fully correct before the higher levels can function at all. This would create an inherent thresholding that could not be avoided by any choice of metric.

To be fair though, the paper's discussion says they are not claiming that models cannot display any emergent abilities, just that some may be mirages.

pulvinar··on Xerox Star 8010 Interfaces, high quality polaroids (1981-)
The Lisa was in development at the time, so apparently someone at Apple shared it. Strange thing to show off in your Xerox demo! (showing how much larger the screen was is my guess)

Lisa, for comparison: https://www.computerhistory.org/collections/catalog/10263089...

pulvinar··on Why is it so hard to pull off a lunar landing?
Folks, NASA's Surveyor program (1966-68) is a better comparison here, not Apollo. It made the first moon landing by the US (after the Russian's Luna 9, 4 months before), and the first landing of the series was successful, though some of the others failed. It was a fully automatic landing, unlike Neil Armstrong's.

It built on lessons learned from the many failures of the Ranger program before it. And the Russians lost 12 landers before their first success.

pulvinar··on ChatGPT for Hardware Design
Except that "copilot" doesn't appear to be registered by itself in this field, only in combinations. But we'll see.
pulvinar··on ChatGPT for Hardware Design
Presumedly they would likewise register "Flux Copilot"?
pulvinar··on ChatGPT for Hardware Design
Update: after asking it to explain absolute maximum ratings and check that for this part (which it knew correctly) and pointing out its error, it finally suggests a reasonable part, the AD5758.

It's clearly missing experience with magic smoke.

pulvinar··on ChatGPT for Hardware Design
Interesting. I tried the same question (each a new conversation, GPT4) and 10V and above it catches, but 8V seems to be in its blind spot. Maybe it hasn't learned anywhere to take absolute maximum ratings seriously?
pulvinar··on There is no A.I.
Fully agree-- thanks for the correction. (still trying to figure out why I wrote that!)
pulvinar··on More than 19k undersea volcanoes discovered
It would be extremely unlikely. Instead, it should be close to equilibrium, with as many eroding away as being created.
pulvinar··on Show HN: Automatic prompt optimizer for LLMs
I'm sure it could be improved, including telling it to do what you suggest. Have you tried it as is though?
pulvinar··on Show HN: Automatic prompt optimizer for LLMs
The prompt it generates looks good and makes sense, but I doubt it's always correct or optimal. But then I doubt a human assistant could give a perfect prompt either (enemy of the good, etc.).
pulvinar··on There is no A.I.
That's most likely due to the way it tokenizes words. You have to be careful here, or else for similar reasons you might imply that dyslexics aren't intelligent.
pulvinar··on There is no A.I.
Say hypothetically that it was a prefect duplicate of a human brain. That would certainly be called a truly revolutionary accomplishment, but that duplicate wouldn't be expected to massively boost productivity any more than adding another human would.
pulvinar··on There is no A.I.
You seem to be saying that to have general intelligence one must be an expert in every field? That rules out quite a number of humans too.
pulvinar··on Apple’s fight against iPhone sideloading was pointless at best, harmful at worst
Yep, serves them right for making desirable products.

Odd that I don't remember anyone calling Nokia anti-competitive when their platform had the market sewn up and Apple and Google had 0%.

pulvinar··on 'Devastating' melt of Greenland, Antarctic ice sheets found
Most people then thought of nuclear winter as more likely.

In the Fort Worth Star Telegram, 27 Sep 1981, is the article "World's mercury inching upward", discussing the science which wouldn't look out of place today. But it wasn't until the end of the 80s that the US government started to take it seriously.

← PreviousPage 7 of 11Next →