Open AI recent drama

[shunsui]

Posting this because there's some important issues here you might want to be aware of.

The last few weeks artificial intelligence company Open AI has been making a lot of news over the firing of CEO Sam Altman by the board of directors and the subsequent return of Altman to Open AI and several board directors stepping down. There have been a lot of good videos explaining a lot of this, but this one is a little more up to date.

https://youtu.be/8e36Y_pdMcw

[Naperville]

Watching the video...

After using GPT-4 for 30+ days straight, I have not logged in to use GPT-4 for a while, maybe 2+ months. I hammered it for info on nuclear war survival, martial arts and health supplements. It was fairly good at the time, better than anything available elsewhere. Saved a lot of time over using the Google search engine.

Yesterday I used Google Bard and it has improved vastly since I last used it.

Bad news that these models are leaking online. It gives our adversaries like China, Russia, Iran and N.Korea a leg up.

[Naperville]

OK, I'm a bit deeper into the video.

Maybe the GPT-4 and Bard systems were not specifically trained to solve the problems that I was asking them, but these are not general AI. There was no genius here, no breakthroughs. NOTHING. Just a regurgitation of web scraped documents formatted to look like one person was behind the responses.

What if these or THIS video is mis- or disinformation and was produced by the CIA to make it sound like the USA has a tool to crack encryption? We all know the media are used by the CIA to spread "fake news" ....do you believe in extraterrestrial life here on earth? That was last months breaking news, er, FAKE NEWS.

I'll believe that global encryption is broken when tens of millions of credit cards have been pawnd and Internet shopping ceases to be a thing we can securely do.

[Naperville]

I suppose there are others here who have had similar experiences but 2000-2003 I was in open meetings in Silicon Valley where cryptographic methods were discussed where people with certain clearances had to get up and leave the room. They could not even be present when the subject was discussed.

Funny thing that I heard one day. Most of the technical books on verboten subjects have forced "errors" in them. Only people cleared by professors to learn the material or who have clearance know what the true vales are in the texts, because they have a right to know the material, and this is done to contain and control the spread of the science or technology.

Take what you will from my experiences and things I've heard what you will.

[Mushroom]

Watching the video...

After using GPT-4 for 30+ days straight, I have not logged in to use GPT-4 for a while, maybe 2+ months. I hammered it for info on nuclear war survival, martial arts and health supplements. It was fairly good at the time, better than anything available elsewhere. Saved a lot of time over using the Google search engine.

Yesterday I used Google Bard and it has improved vastly since I last used it.

Bad news that these models are leaking online. It gives our adversaries like China, Russia, Iran and N.Korea a leg up.

You use Ai for “research?” Do you trust that the information Ai provides you is always correct? I would strongly caution against using Ai as a reliable source for accurate information. (And absolutely never as a source of authority.)

[Naperville]

Watching the video...

After using GPT-4 for 30+ days straight, I have not logged in to use GPT-4 for a while, maybe 2+ months. I hammered it for info on nuclear war survival, martial arts and health supplements. It was fairly good at the time, better than anything available elsewhere. Saved a lot of time over using the Google search engine.

Yesterday I used Google Bard and it has improved vastly since I last used it.

Bad news that these models are leaking online. It gives our adversaries like China, Russia, Iran and N.Korea a leg up.

You use Ai for “research?” Do you trust that the information Ai provides you is always correct? I would strongly caution against using Ai as a reliable source for accurate information. (And absolutely never as a source of authority.)

What I did was cut and paste all of the responses that I received. Then I checked some of it. None of it was important, I just wanted to see how it's answers changed as it advanced over the years.