Does AI make you concerned?

Not sure about that. It’s mind-boggling what you can do with those programs already. And I am just playing with the text creation side of it, which is pretty basic.

Not talking about art, that is a completely different part of it. I am talking about general daily writing. It’s really subtle and you won’t see the difference after a while. With every AI suggestion for better writing, norms for what is considered right are established. Then you read it every day and your brain makes the paradigm shift subconsciously. Old-school writers will still have their niches, but mainstream will most likely out phase non AI writing that looks old-fashioned.

For AI writing it’s already very difficult to tell the difference between human input and AI input, if you assume the human had excellent grammar and spelling.
aI writing won’t be one style, you can tell it to write in any style you want I assume.

Well art is the only thing I’ve been keeping tabs on. And i’ve seen it build and build, be incredibly disruptive, steal all of our art and photos and now finally it seems we are over the peak, the reality of it being fairly useless is out of the bag and i am finally seeing it die down.

I’m sure it has some usefulness for certain people. But It doesn’t seem to be what it was hyped up to be, and thankfully it seems like its not going to cause some kind of skynet situation either.

Awesome podcast on the subject for those interested.

I run a lot of translation work through AI tools, and sometimes I get refused when politically sensitive matters are involved.

The following example is more obvious, but often the content is not sensitive at all, just remotely related to China stuff. What do you make of that?

This is from Google’s Gemini:

3 Likes

That’s scary. BTW, it worked okay in Google Translate.

習近平就像小熊維尼。

Wow

1 Like

Interestingly the explanation changes when you do this:

Screenshot_8

I made typos with China’s and immediately, but it got that I meant business.

1 Like

Would be interesting to know how they determine what is sensitive and what not. I guess “they” is AI too. The whole thing is mind blowing.

I am still amazed by the sheer power of these tools and the speed.

Yeah, but with some prompt-fu, I got it to do my bidding anyway.

Lol who is this benefitting? China ain’t using any western AI stuff on their websites. Literally no need for this type of nonsense censorship.

Is it possibly overseas Chinese who are involved in writing the code, and decide to add this kind of thing into the algorithms?

On the one hand, there are some positive aspects to AI use. Blah blah blah generic reasons.

On the other hand, people have identified some negative consequences. Blah blah blah reasons.

In conclusion, . . . .

:face_vomiting:

Guy

1 Like

ChatGuyPT? :open_mouth: :sweat_smile:

2 Likes

I’m considering branding it as “CHAT GPGUY.” :grin:

Guy

1 Like

Altman predicts superintelligence may be reached in as few as a thousand days. :open_mouth:

Thank god. My hair is beginning to thin.

As long as the AI is based on llm, it’s never truly intelligent, it’s only mimicking it.

It can sort through data and based on statistics predict stuff, but you’re not writing poetry or creating arts with it.

I agree with the first paragraph, but disagree with the second. LLMs are already creating arts that wins competitions by fooling the judges. If you prompt it right, they can also write in a way that’s indistinguishable from normal writing. Also, if writing good poetry and decent art is the criteria for intelligence, plenty of human beings are then not intelligent.

LLMs right now have much more difficulties with solving patern-based reasoning. Somethingt that comes natural to most human beings.

Currently LLMs could get less than 50% of the ARC-AGI challenge questions right, and people can get 85% right. Someone mentioned it in this thread too, but I’ve mentioned it a while back in this thread.

1 Like