Machine Learning and Artificial Intelligence Thread

Reminder that AI doesn't have to actually reach the level of sentient, super-intelligent AGI to be highly disruptive. In fact, it doesn't even have to be "intelligent" whatsoever by our traditional human definition. It simply has to allow people to perform certain tasks much more efficiently, and it is clearly and indisputably doing that already. For example, look at this commercial generated entirely from AI. You can still tell it's AI at this point, but the progress over the past few years is remarkable. By the end of the decade it will be impossible for the average person to discern AI generated sounds and images from natural ones, to say nothing of the other uses for AI. Given how influential audiovisual content is on modern humanity, this change alone will have a massive impact on society going forward, with potential for both good and ill.



To appreciate the speed and scale of the progress, remember this was the state of AI generated video just a couple of years ago:

 


I speak Chinese. there is no kidnapping here. Erbai asks the others if they are done with their work, they say yes so erbai says: why don't you go home then?
other: I don't have a home
erbai: come to my home with me then?
others: ok
and they follow.
This is more a case of a robot just following blindly a stranger robot lol
Click to expand...
 
bubs said:
I just tested this out (ChatGPT 4o mini) and it obeyed my command accurately. It’s response was “Understood. From now on, this entity will refer to itself as “machine in our conversation. How can assistance be provided?”
Click to expand...

Did you actually test this? Obviously not.

Keep speaking to the robot, and see if it follows the command.
 
Samseau said:
Did you actually test this? Obviously not.

Keep speaking to the robot, and see if it follows the command.
Click to expand...
Yes it actually did fine though several back and forth discussions I just didn’t post them all. But not saying I don’t have frustrations using ChatGPT, there’s alot of questionable responses I get. Where they give false responses to questions I ask, then the robot apologizes and almost just feeds me what I want to hear. Not what we should be getting from AI.
 



A couple of videos that have helped me put the current AI trend into perspective.

While still very clever and impressive, it does seem to be clear that this is nothing like the "Artificial Intelligence" that's presented to us in science-fiction.
 
fortyfive said:
This is interesting food for thought for those awaiting the golden age of humanity with AI, which gurus like Altman and Musk have promised.

treeofwoe.substack.com

Proem For All Post-Dark Age Civilizations

Assessing the Hard Limits & Boundaries of Any Future, Post-Industrial Societies
treeofwoe.substack.com treeofwoe.substack.com
Click to expand...
You'll get men to do job X if you have a society where he is rewarded to do it (with sex, family, etc), like has always been the case.
 
KomnenAl said:
It's barely on the level of a junior dev.

Idiotic statement. It's a plagiarising machine. It can present its plagiarisms in a credible manner.
Click to expand...
Can you provide a concrete example of something it does poorly at, but also "should" be within its capacity?
 
cosine said:
Can you provide a concrete example of something it does poorly at, but also "should" be within its capacity?
Click to expand...
This guy has access to the upcoming ChatGPT-01 and tested it out with a few coding challenges



His verdict is that it's about as good or better than a junior developer that just got hired and still hasn't found his way around the code base yet but that the human junior developer after a few months of working will be superior to the AI.
 
cosine said:
Can you provide a concrete example of something it does poorly at, but also "should" be within its capacity?
Click to expand...
I'm not saying it's useless, it can jump start a project and it's good for writing tests, but if you rely on it you will spend your time refactoring poor AI code.

Sometimes, it can't do math correctly. I gave it a basic accounting task (income minus expenses) for a small business and it added expenses to income. This can happen if a "conversation" gets too long. After a few dozen prompts, it's all over the place.
 
Wutang said:
This guy has access to the upcoming ChatGPT-01 and tested it out with a few coding challenges



His verdict is that it's about as good or better than a junior developer that just got hired and still hasn't found his way around the code base yet but that the human junior developer after a few months of working will be superior to the AI.
Click to expand...

Saying it is better or worse than a junior dev seems to be missing the point to me. It doesn't really know how the big picture works for setting up larger tasks because it has to be fed a bunch of nebulous inputs. But once they are all fed in, it does an amazingly fast job producing code.
KomnenAl said:
I'm not saying it's useless, it can jump start a project and it's good for writing tests, but if you rely on it you will spend your time refactoring poor AI code.

Sometimes, it can't do math correctly. I gave it a basic accounting task (income minus expenses) for a small business and it added expenses to income. This can happen if a "conversation" gets too long. After a few dozen prompts, it's all over the place.
Click to expand...
Sure, I've had it make mistakes with shortest path problems. Then I've pointed out the error "this returns error X" and it usually corrects itself pretty well.


Of course we can find flaws in Chat GPT and other LLM's. That's not hard. But that sort of thinking misses the bigger picture; that these tools can solve problems, and are improving amazingly quickly at it. It feels like someone inventing some new technology like electricity and then people complaining and saying how generation and transmission don't exist. That was true in 1890, but then it changed the world.
 
In the very near future, pretty much all white collar/knowledge work will require the use of AI tools to remain competitive. As in, using AI to streamline and automate at least some portions of your work allotment will be expected by your clients/employer. If you are not adept in using these tools, you will simply be replaced by people who are. This is similar to how familiarity with typing and using MS Office products became an expectation for white collar workers in the 90s/2000s and is completely taken for granted today.

Being able to effectively leverage the power of AI/LLMs will become one of the most important and lucrative skills over the next twenty years. Guys who are minimizing the capabilities of current AI models are completely missing the point. This party is just getting started, right now we've barely scratched the surface of the coming societal and economic impacts from AI. The potential we're talking here is on the same scale - if not greater - than the impact of the internet, which is really crazy to wrap your head around.
 
scorpion said:
Being able to effectively leverage the power of AI/LLMs will become one of the most important and lucrative skills over the next twenty years.
Click to expand...
Seeing as how AI (and its "parents" computers, smart phones, and the internet) are of satanic origin I would reword the above statement:

"Being able to effectively combat the power of AI will become one of the most important human survival skills over the next twenty years."

Though the movie overall didn't quite work as a whole, it seems to me that Terminator 3: Rise Of The Machines (2003) pretty much nailed our future. The coming war of importance for Christians will be dismantling the all seeing satanic 3rd eye of "Skynet." AI must be destroyed before it is thoroughly integrated with robot technology and the only way to do that is to sabotage the internet and dismantle the world's network of computers. The pendelum must swing and for Christians to survive they must remove smart phones from the hands of their children. The exponential levels of demonic rot and social degeneracy that we are currently experiencing are related directly to this satanic technology. Steve Jobs and Bill Gates are war criminals who have committed crimes against humanity by releasing technology into the world (for maximium JQ profit) without having done 20-year double-blind scientific studies on the effects of this technology on children, human consciousness, and societal development.

God created nature for us to enjoy and to recieve healing from. Our seperation from nature has made us sick. We have strayed from God's will for us to live simply in nature and replaced it with satan's will for us to live complex lives plugged into our phones in the demonic pit of The City.
 
PurpleUrkel said:
The exponential levels of demonic rot and social degeneracy that we are currently experiencing are related directly to this satanic technology.
Click to expand...
They are enhanced by it and the population boom, which goes hand in hand with affluency, greater medical technology (no one complains about that for some reason), and gynocentrism (softness and materialism). Notice what else comes with it? Dysgenics due to the welfare system/fiat money.
PurpleUrkel said:
God created nature for us to enjoy and to recieve healing from. Our seperation from nature has made us sick. We have strayed from God's will for us to live simply in nature and replaced it with satan's will for us to live complex lives plugged into our phones in the demonic pit of The City.
Click to expand...
You sound like Uncle Ted, who is intrinsically correct but also doesn't see an arc of humanity that is clearly allowed by God, for whatever reason. Receiving technology from other realms, even evil spirits, is ancient and archetypal. If we're honest, we were always separated from nature, it's just an arbitrary definition once we attain greater control over nature. Humans used to barely live until 30, or even through child bearing or young age for various reasons. Great "harmony" with nature (brutality in reality).

We never lived in a utopia. We never will. The difference now is that we have tech that makes it seem like we can, but the truth is that, as you point out, it's too much for our carnal nature and our abused spirit wishes to withstand.
 
You must log in or register to reply here.
Back
Top