[Sliced]

We now know that AI can be used to make videos, because it can generate mass photos. hence videos in longer format of pictures.

Is THIS the end of the news industry? no one will be able to tell fake news from real news.

[Cuthbert]

No, it's not, and informed readers will still be able to tell. The prose of AI often lacks character and can be full of inexactitudes. Reporters still have to go out and gather the information. Furthermore, stories worth reading are usually written by journalists who identify themselves.

[Lindros]

No, it's not, and informed readers will still be able to tell. The prose of AI often lacks character and can be full of inexactitudes. Reporters still have to go out and gather the information. Furthermore, stories worth reading are usually written by journalists who identify themselves.

This is true, for now.

Honestly as a programmer, I said to myself "ChatGPT is clearly just a tool, it's helpful but very limited. There are likely limitations to how good it can become." I now believe this was honestly a cope.

GPT-4 is said to be coming soon and 500x more powerful than GPT-3. We really don't know how quickly this stuff can and will likely advance and to what level.

If they can automate complicated programming problems, they're going to be able to automate contract legal defense, designs of buildings, video production, flight responsibility's typically designated for pilots, and much, much more.

We are approaching a singularity, likely, of unparalleled, exponential improvement to these model's abilities and performances. Humans have so thoroughly documented their thoughts online for long enough to train these models to be able to largely replicate, in the future, what the average person can say, do or create.

[Cuthbert]

This is true, for now.

Honestly as a programmer, I said to myself "ChatGPT is clearly just a tool, it's helpful but very limited. There are likely limitations to how good it can become." I now believe this was honestly a cope.

GPT-4 is said to be coming soon and 500x more powerful than GPT-3. We really don't know how quickly this stuff can and will likely advance and to what level.

If they can automate complicated programming problems, they're going to be able to automate contract legal defense, designs of buildings, video production, flight responsibility's typically designated for pilots, and much, much more.

We are approaching a singularity, likely, of unparalleled, exponential improvement to these model's abilities and performances. Humans have so thoroughly documented their thoughts online for long enough to train these models to be able to largely replicate, in the future, what the average person can say, do or create.

Let me know when it can get more people to Church. From what I've seen it's being shaped to be very politically correct in its answers. Perhaps the more private versions of it has greater usages, but I'm just not too convinced by it all. It'll still take a programmer to use the code it spits out, and a skilled reader to make use of its scribblings. It's all just zeros and ones on a little silicon chip made by imperfect humans at the end of the day.

[Lindros]

Let me know when it can get more people to Church. From what I've seen it's being shaped to be very politically correct in its answers. Perhaps the more private versions of it has greater usages, but I'm just not too convinced by it all. It'll still take a programmer to use the code it spits out, and a skilled reader to make use of its scribblings. It's all just zeros and ones on a little silicon chip made by imperfect humans at the end of the day.

I said that it will get better, not that it will be used in beneficial of benevolent ways.

[Cuthbert]

I said that it will get better, not that it will be used in beneficial of benevolent ways.

I think it will be used malevolently, I can already see it being used as a little "Fact" Checker, ramping up the left-wing propaganda all over the place. I hope it does get better for people who need it, and doesn't take too many jobs from those who deserve them. I asked it about marijuana and it contradicted itself by going something like, "there's no evidence it's harmful [...] some who use it experience psychotic symptoms." Don't even get it started on religion. It's very indecisive and seems to regurgitate what's popular. I suspect for many Atheists it's just another futile attempt at replacing God.

[Omniscient]

If they can automate complicated programming problems, they're going to be able to automate contract legal defense, designs of buildings, video production, flight responsibility's typically designated for pilots, and much, much more.

The first legal trial based on AI incoming. That's gonna be crazy when an AI Judge is going to decide a case based on evidence input by the Defense and Prosecution.

[Lindros]

The first legal trial based on AI incoming. That's gonna be crazy when an AI Judge is going to decide a case based on evidence input by the Defense and Prosecution.

They're going to train it on a bunch of BS woke and biased, curated data because of "AI ethics"..

Oh, unfortunately GPT-Judge has determined you made an insensitive comment on Facebook in 2009. Due to this, when you defended yourself against a Haitian criminal who broke into your house to rob you, you will now be charged with a hate crime and summarily executed.

[Omniscient]

They're going to train it on a bunch of BS woke and biased, curated data because of "AI ethics"..

Oh, unfortunately GPT-Judge has determined you made an insensitive comment on Facebook in 2009. Due to this, when you defended yourself against a Haitian criminal who broke into your house to rob you, you will now be charged with a hate crime and summarily executed.

I fear you're correct on that based on racist questions I've already asked ChatGPT. It's obviously programmed with bias, which is crazy considering the AI should be based off facts not feelings. I won't share the questions I asked but the answers it gave were annoying, boilerplate, and inaccurate. Responses were basically "that's a mean question, all people are the same".

[Baby]

If AI is coded & designed, bringing it back to 0 & 1 wouldn't any form of AI almost have to be biased in any form or fashion to produce a result? I'd prefer not to have any computer sorting us out. Life's hard enough as it is with flawed judgment from human input.

[Sliced]

If AI is coded & designed, bringing it back to 0 & 1 wouldn't any form of AI almost have to be biased in any form or fashion to produce a result? I'd prefer not to have any computer sorting us out. Life's hard enough as it is with flawed judgment from human input.

Yes but what really is a BIAS result? every action you take is an action someone else took that want the same result that you wanted right?

[Baby]

Yes but what really is a BIAS result? every action you take is an action someone else took that want the same result that you wanted right?

It's full circle & unexplained at the same time lol. Very true.

[OMSPK]

We all know how humans are. I feel as though the bigger and more "interactive" (for lack of a better word) the AI becomes, the use-cases will also grow to where we start to see laws being produced by politicians to stop the AI from being able to judge, moderate, or determine certain aspects of our existence. I don't expect this to happen for years and years to come, however I do think it could happen in the near future as we progress.

[Cuthbert]

It's obviously programmed with bias, which is crazy considering the AI should be based off facts not feelings. I won't share the questions I asked but the answers it gave were annoying, boilerplate, and inaccurate. Responses were basically "that's a mean question, all people are the same".

Yeah, this will probably be why it's still limited, making sure it's PC enough. How will HF, and other forums, combat GPT-generated posts?

[Nate]

I’ve been seeing people use like AI chat bots to finish homework assignments and some even get it to code malicious shit