The ARTificial Onslaught

Bad news for programmers?

Only for the boring, repetitive stuff.

Society still needs to have that conversation, which was deferred back when Jimmy Carter was President, on how society will cope with jobs being automated out of existence.

There’s a finite amount of work which needs to be done each day so as to keep the trains and planes running, the world clothed and fed — if a not-insignificant portion of that can be automated how should society react?

Reducing the work-week was was done during The Great Depression to increase the number of jobs for workers would be one approach (and yes, I capitalized all 3 letters there — do folks still have stories of what that time was like for their families which are passed down? My family does, and if folks don’t have such traditions, it explains a lot about what is currently wrong with the world).

2 Likes

Yes. The story of my grandmother jumping the line that snaked around the block and climbing up a fire escape in a local North Philadelphia textile mill duing the Great Depression (not really sure what was so great about it) and talking the manager into giving her a job as a seamstress.

The young ones in our family can recite that story before they can speak. :flushed:

1 Like

Not a CHANCE that your OneDrive data is ar risk:

2 Likes

99.999999% chance of human extinction by AI?

So says this AI expert, but Musk pooh pooh’s it as only 10-20% chance…

1 Like

So, we’ve got that to look forward to…

2 Likes

I get those odds every day driving on the New Jersey Turnpike.

I ventured there ONE TIME in my life - NEVER AGAIN!

1 Like

This is a really good update and summation of the current status of AI-generated video. It sure has come a long way in just one year! Scary in its implications, awesome in its potential.

Marques Brownlee does a great job as usual. Thoughtful, insightful. Remember: “This is the worst this technology will ever be from here on out.”

Man, that 100% AI-generated golden retriever pups clip is astounding. So are the other good examples.

2 Likes

I get Musk’s hyperbole, but Jaime Dimon on the same page?

WSJ - Musk and Dimon Predict AI Will Soon Be Smarter than any Human

Inventor-invention personal bias. At the end of the day, AI is man-made and, being man-made, it will have flaws making it, in its best-case scenario and in its best light, no better than the humans who made it. Such flaws like their creators’ flaws open it to vulnerability no matter how intelligent its creators in their hubris think it may be.

Clearly AI will be able to do many things better than sentient beings but I don’t think it’ll ever be “smarter”.

2 Likes

Exactly. Machines are rather well-suited at making short work of repetitive logical processes, but you cannot never imbue intelligence or out-of-the-box thinking in them.

Ok - couldn’t stay on the sidelines anymore - don’t bet on your conclusion. The very anger/frustration in the early hallucinations more than a year ago border on irrational behavior, not just confusion. It will take longer than yer end, but it is scary close…

The “irrational” behavior in AI is noise, from a seed of some sort or another. The whole premise of AI is pattern matching with noise applied for a random factor. That’s why AI isn’t alive or real or intelligent in the slightest. It is something that is acted upon, and not something that acts of its own volition. Its illogical responses are the result of a mistaken premise: that the mere act of adding randomness makes something intelligent and feeling. It is like adding an extra part “just because” that can go out at any time. AI won’t go on a rampage unless people put the more volatile deep learning variety into automated processes that do not need that degree of randomness. And while Elon Musk believes it can act on its own, I strongly disagree on that point. In reality, what happens is a random seed falls into a logic path that shouldn’t be there. AI is quite logical if you can unwind it. It is when people overthink it that it seemingly becomes some sentient being which it is not.

4 Likes

Which just underscores how incredibly difficult a true General AI is, and how far we are from truly Self-Aware hardware/software.

Three interesting fictional takes are:

  • Heinlein’s The Moon is a Harsh Mistress — this just posits that a computer will “wake up” and become self aware after exceeding a certain threshold of connections — as is a neuron was equivalent to a transistor — charmingly naÏve
  • The Cybernetic Samurai — one of the characters in it has as their favourite novel the afore-mentioned work — annoyingly, it gets one aspect of computer science wrong, positing algorithms which have random results, but then replacing them with a sensor of atomic decay for a truly random result (these should have been reversed to my mind), but von Neumann’s quote is still relevant:

Anyone who attempts to generate random numbers by deterministic means is, of course, living in a state of sin.

  • The Turing Option — marred by the AI researcher’s association with a convicted felon, this book does still manage to convey something of the difficulties involved in actually creating general purpose artificial intelligence
5 Likes

I think I’ll take my overthinking elsewhere…

1 Like

I think the main element missing is rumination and adaptation. If they allow information (text, images, sound, whatever) to be processed on an ongoing basis to a system that has a goal, and possibly allowing tinkering model parameters ‘positively’ when then goal is getting closer, wouldn’t we be approaching what the brain does? I’m very curious where we’ll be in a few years.

1 Like

That was super interesting. The chat with the NYT reporter where “Sydney” suggested he leave his wife. :joy: As for whether this is all math and noise, as long as nobody can figure out what the heck consciousness is (as in “what is the physical meaning of a vantage point”) I don’t think there’s anyone who can confidently take the high road on “intelligence” in complex systems.

That therein specifies lack of autonomy and all logic paths unwind back to a seed that is wholly controlled by process, not personality.