Fear is a bad advisor. I think you're not to blame for anything, and you haven't failed either. AI is here and is being expanded and improved; it's faster and cheaper than humans, plus it's a huge business, and progress can't be stopped. It's much more important to be sufficiently informed about the advantages and disadvantages of AI and how to protect yourself from its side effects.

Young people should be empowered to think critically and question their own actions from time to time. Education will NOT come from AI manufacturers, because they consistently promote all their products positively because they want to sell their product.

Do you remember horse-drawn carriages as a means of transporting people and goods and then the big shift to single-engine cars? In the beginning, there was great chaos, many injuries, and few rules. Legislators then established a few rules and introduced mandatory driving licenses. Can cars be reversed? No, and the same will happen with PCs, AI, and robots. We must ensure that rules are established and adhered to.

Everything we have invented, plus nature with its plants, animals, and weather, always has two sides, depending on whether they are struck by lightning or fall victim to a flood. On the other hand, nature can be very beautiful; we have to adapt and learn.

If we don't fully utilize our brains, we will perish. So let's learn about the dangers and the proper use of AI.
There will always be criminals who will and want to deceive us, and only the rule of law and self-protection can help against them.

 

Ok. Sorry, it wasnt obvious. But it is a political choice. What will happen should be fought, not accepted.

It is.

 

better to bash coconuts together and catch fish in your hut next to your kratom field.

 

"Maybe we deserve it"...the we is a first person plural pronoun and is inclusive of all of humanity whereas your use of "they" implies that you've stepped away at some distance from the human race and are outside of the circle of man! Are you of an alien race and coming out of the interstellar closet to us or is there another less space-ific explanation?

 

Well, OpenAI launched GPT5 and it included this informative graph about it's supposed performance:



I'm not going to take them particularly seriously. Massive tech hype as usual, but as far as misinformation goes this is just atrocious. And this is from a company that's valued at half a billion dollars which makes 5 billion in revenue (not profit).

 

I have absolutely no idea what that graphic implies. It seems the work of a 7 year old.

Edit...what I meant to say.

 

It shows the revised numbers for the amount of jobs created last month.

Actually, I have no problem believing those numbers. The first 2 columns on the right show first-pass coding accuracy, during periods of time where their numbers were lumped together in 1 category instead of breaking them out separately.

The GPT 5 model shows the baseline first pass accuracy with raw output from GPT5. It is showing approximately a 5.8% increase in total accuracy over the OpenAI v3 numbers. It breaks those into 2 categories, one with human intervention and thinking, and the other category without human input. So they are actually showing a 22.1% increase in first pass accuracy with human intervention, over the results without human intervention.

 

Thanks for your interpretation.

Translation in street jargon: AI software is getting better every day and is being used by more and more people.

 

Somehow that graph doesn't quite convey the numbers properly. o3 is twice performant than GPT4 but hmh, the graph somehow distorts that too. And visually that certainly looks nothing like the ~5 percentage unit boost.

Since graphs are meant to deliver information visually, why in the sweet fuck did they do this? Wilful misinformation or sheer incompetence? You'd imagine a half trillion business could hire a copy editor to check this stuff.

 

The reason why the two columns to the right scale to the percentages they do, which you are referring to as distortion; I agree. It is not proportional to the percentages. Of course, that skew is probably to make the new model's numbers look like a more drastic improvement than the numbers suggest. Or that was done to minimize the importance of the two columns before they began measuring these data. Usually that kind of thing is done to get all the data/chart to all fit on one page, which is obviously not the case here.

All I would even try to take from it is that it is a 5% overall improvement over prior best performing model, and that there is a 22% improvement when the GPT output was coupled with actual "thinking". From there, we can probably expect future models numbers to gradually reduce the gap between raw output, and monitored output.

 

Ai blows smoke up the ass of crazy people and confirms pre existing biast to the most extreme degrees, other than when people use it to agrigate information, if i see someone who's has a "relationship" with their ai you can tell what ever crazy stuff they've come up with is what they talk about with it all the time (aka program it to talk about)not proof of the crazy thing they trying to find out

 

Honestly, I guess I’ve just been lucky so far. In the last few months, AI has helped me level up my knowledge in all sorts of areas—science, music, art, politics, you name it. Whenever I wanted info on something, I could get it super fast.
With computer and network tech in particular, I've never been able to make progress this quickly before; AI just makes everything way easier. From what I can tell, AI has only brought upsides for me and what I know.
Same goes for my wife—she’s moving forward way faster in her job and can get her project work done quicker and better.
As for music, I barely use AI, to be honest.