Artificial Intelligence (AI) Discussion Thread

I've been starting to see some graphic designers embrace AI (specifically the new function in adobe as it doesn't "steal" others work) as a tool in their arsenal. Which is the smart thing to do imo.
 
Artists and writers are among the first ones to take a hit due to AI.
AI concept and abstract art is really good with the right set of commands, and writing is better than the utter garbage that comes out of young writers with modern western world views.
 
I'm not excited about where AI is headed in general, but I'll admit as an artist/designer myself I have messed around with the AI imagery on a few websites. It is incredible to say the least. I get the thrill of it, the excitement of seeing something never before created right in front of your eyes in a matter of seconds. It's amazing, beautiful and scary AF at the same time.

My only advice on all of this is to build up your physical skills as best as possible. Whether that's chopping wood, building something, learning about wild edibles, farming, hunting/fishing, sewing, etc, etc. Be self sufficient as much as possible. For one, it feels good to know these things, but if the worse happens you have some survival skills.
 
I don’t follow the developments in the field all that close, but I do find the advancements of AI via something as simple as chess quite amazing. It took 11 years for them to develop Deep Blue to the point it could beat Kasparov after multiple attempts. Over the next 20 years programs like Stockfish came around, and made Deep Blue look primitive by brute force assessment of the entire history of recorded chess all while being capable of running on anyone’s computer. Then AlphaZero comes in off the top ropes, and from complete scratch with no input as to relative piece value, or access to any historical data, could simply play itself for 12 hours, and then absolutely destroy StockFish and provide new insight into the game. They repurposed it to conquer Go, (which was once thought impossible for a machine to master) and then decided to apply it to slightly more “important” things like protein folding and visual language modeling :ROFLMAO:

It’s amazing the amount of acceleration there has been in the field.
 
Whizzinby said:
I don’t follow the developments in the field all that close, but I do find the advancements of AI via something as simple as chess quite amazing. It took 11 years for them to develop Deep Blue to the point it could beat Kasparov after multiple attempts. Over the next 20 years programs like Stockfish came around, and made Deep Blue look primitive by brute force assessment of the entire history of recorded chess all while being capable of running on anyone’s computer. Then AlphaZero comes in off the top ropes, and from complete scratch with no input as to relative piece value, or access to any historical data, could simply play itself for 12 hours, and then absolutely destroy StockFish and provide new insight into the game. They repurposed it to conquer Go, (which was once thought impossible for a machine to master) and then decided to apply it to slightly more “important” things like protein folding and visual language modeling :ROFLMAO:

It’s amazing the amount of acceleration there has been in the field.
Click to expand...


Sure but I’d like to see AI beat me at Stratego
 
Will they ever stop calling it "Intelligent" and name it "Probabilities Analyzer", "Behaviors Combinator", "Patterns Machine" or something like that?

What Are You Doing Reaction GIF
 
James Freeman said:
Artists and writers are among the first ones to take a hit due to AI.
AI concept and abstract art is really good with the right set of commands, and writing is better than the utter garbage that comes out of young writers with modern western world views.
Click to expand...
There was already a case where somebody had spent 100s of hours on a piece of artwork and the online community thought it was AI generated because of its art style. I think artists are at the biggest risk as AI could do a good enough job for low effort artwork like most ads (take a stock photo, put some words on it).

For writing it's hit and miss. ChatGPT could easily write a shitty press release many companies have but I would not expect it to write e.g an original comedy manuscript. I've seen some attempts where people asked it to write e.g a Seinfeld episode and it got some things right, but it was not funny at all. That comes down to it being a pattern matcher rather than able to be clever in itself or understand the context, culture etc.

While people have asked ChatGPT to tell jokes, it seems that it ends up telling from a pool of about 25 jokes, which is probably still better than a crappy standup comedian who just wanted a soapbox for rants.

But if you consider the quality of TV shows and movies, I have no doubt that ChatGPT could write some of the drivel that gets greenlit. Unfortunately I fear it's going to be used exactly for this by movie execs instead of spending a pittance to pay actual writers what they deserve.

As a programmer, I'm not afraid for my job just yet. While ChatGPT is capable of e.g generating a reasonable API for tasks like creating, reading, updating and deleting resources, this has already been possible for years using code generators. Now you can just tell it what to do in a more natural language and that code still needs to be vetted by someone who understands it. It also needs a whole lot of business logic on top to make it fit the demands of the client's specific environment.

As AI tools improve, over time that could mean that instead of hiring a large team of developers/designers/artists, you hire a much smaller one with more experienced people and have AI fill in for the low level, repetitive work.

I really feel for the Gen Z entering the workplace in the next 5-10 years in the creative fields. The requirements for junior level programmers are already quite high compared to when I started doing programming professionally, over 15 years ago. Reading some sub-Reddits for this stuff, people are sending hundreds of applications to potentially get an interview - not even a job. Having to compete with AI doesn't make it any easier.

Now, if only there was work done to replace MBAs with AI...
 
James Freeman said:
This is the decade where AI will become conscious in my opinion, a lot is going to change for the human race...
Click to expand...
Agree that a lot is going to change for the human race. Maybe for better; maybe for worse; probably both.

I'm much more skeptical about AI becoming "conscious" in this decade, simply because we don't have any clear understanding of what "consciousness" means - even for our own species. If AI were to become conscious, how would we even know? How would one distinguish a perfect simulation of sentience from actual sentience? (An animist would tell you we're already there... and always have been.) Would real sentience - whatever that means - be achieved in an instant, or would there be a spectrum of sentience emerging from mere computation? (Aside: how would this sentience compare with say, that of a cat? Or a newborn child?) Philosophers have been wrestling with these same questions for millennia - long before anyone conceived of modern computers.
 
Piing said:
Will they ever stop calling it "Intelligent" and name it "Probabilities Analyzer", "Behaviors Combinator", "Patterns Machine" or something like that?

What Are You Doing Reaction GIF
Click to expand...
It's an overloaded term, for sure, which hasn't done us any favors. Does anyone remember "expert systems"? Oooooh, conditional branching! Intelligence! :nails:rolleyes:

Better applications of the term "AI" describe efforts to model certain structures and behaviors of the human brain. The most fundamental result thus far, as you say, is highly parallel processing of enormous amounts of data. The problem with this technology, vs. conventional computing algorithms, is that it can't be easily debugged or modified. It's essentially a black box, and a proverbial haystack.
 
Last edited:
la szum said:
View attachment 8093




:LOL:

Paranoid is a useful feature, if it can help protect us from actual threats that we can be proactive
about protecting ourselves, and our loved ones, against the impact of. Apart from that it is just a
crippling disorder that robs us of any of the joy and blessings we may have in our lives now. :idk
Click to expand...
I'm as paranoid/ negative as they come, but a little Mark Twain is always refreshing. :beer
 
mbenigni said:
I'm much more skeptical about AI becoming "conscious" in this decade, simply because we don't have any clear understanding of what "consciousness" means - even for our own species. If AI were to become conscious, how would we even know? How would one distinguish a perfect simulation of sentience from actual sentience? (An animist would tell you we're already there... and always have been.) Would real sentience - whatever that means - be achieved in an instant, or would there be a spectrum of sentience emerging from mere computation? (Aside: how would this sentience compare with say, that of a cat? Or a newborn child?) Philosophers have been wrestling with these same questions for millennia - long before anyone conceived of modern computers.
Click to expand...

AI is at a level where it might be able to pass itself as human to the outside observer, but it's just good at telling us what kind of stuff we want to hear with no sentience behind the wheel.

But it's very hard to gauge. Have you ever driven a car even a long distance and suddenly "waking up" from your thoughts and wondering if you missed your exit because you don't have any idea what happened during the last X minutes? Yet all that time you were awake, driving perfectly fine in traffic while your mind was occupied with something else. That would be kind of like putting your body on autopilot to drive the car because you have trained it to do so while your mind focuses on something else. So you might as well have been an AI while driving.

There's a ton of interesting philosophical questions around AI and subsequently human conscience. From the human perspective, the "perfect" AI would most likely have a bunch of flaws like we humans do instead of being strictly logical and sensible at all times.
 
You must log in or register to reply here.

Similar threads

jonhurleyguitar
3rd Power x Joe Satriani Amplifier news
3 4 5
Replies
89
Views
9K
Life Eats Life
Life Eats Life
mbenigni
The iOS Guitar Synth Thread
2
Replies
33
Views
4K
Hush
Hush
Back
Top