Elon Musk to develop 'TruthGPT' as he warns about 'civilizational destruction' from AI

a reply to: ArMaP

They do exist I had a good think about it and it's everywhere although basic. My local taxi rank uses one that remembers numbers and gives suggestions on pick up points and destinations.

I wasn't thinking decisions, that's a well established line in the sand for the most part. That said the various systems are getting good, the French government can attest to that with systems like this: Link - AI assists French government with swimming pool taxes

I don't think real intelligence from an artificial system is that near.

Totally agree although they'd needn't be to totally revolutionise society, we're in the midst of a technological revolution. I'm not that technical but integrated systems seem to be the future which makes me think that privacy rights and laws will need working on. I'd say courts, laws and society is usually a few years behind in that regard.

originally posted by: RAY1990
They do exist I had a good think about it and it's everywhere although basic. My local taxi rank uses one that remembers numbers and gives suggestions on pick up points and destinations.

I'm a programmer and I work on a software house.
One of our main programs uses keywords to help find the information the user is looking for, to allow people to search expressions that are not really on that specific record's text.
One example would be "ship". The text may have the word "frigate", but most people do not search for such specific expressions, so the keyword "ship" is added to that record.
In one recent upgrade I added the possibility of the program suggesting keywords, based on the text of all other records and the associated keywords. It works only by using statistics, no need for AI to do that.

In the same way, many things that look like AI for those that do not know how things work are based on statistical analysis.

Even some AI, like chatGPT, is based (in part) on something similar to statistical analysis

I wasn't thinking decisions, that's a well established line in the sand for the most part. That said the various systems are getting good, the French government can attest to that with systems like this: Link - AI assists French government with swimming pool taxes

I remember that.

Yes, that system uses image recognition, which is somewhat similar to a language model, as both are based on data sets created by humans to feed the AI, so the AI can learn from them.

Totally agree although they'd needn't be to totally revolutionise society, we're in the midst of a technological revolution. I'm not that technical but integrated systems seem to be the future which makes me think that privacy rights and laws will need working on. I'd say courts, laws and society is usually a few years behind in that regard.

That's why some places, like the EU, have been considering AI related laws for the last couple of years or so.

Elon, if you observe him, he's in his own spaceX universe, not on planet Earth.

futurism.com...

Elon Musk Says He's Building a "Maximum Truth-Seeking AI"

Will TruthGPT be able to kill the "woke mind virus?"

After insinuating that he thinks the bot is "woke" and teasing his own "anti-woke" AI, Musk has finally revealed that he's working on "TruthGPT," a brazenly named alternative to the popular chatbot that will serve as a "maximum truth-seeking AI," in an interview with Tucker Carlson on Fox News.

Just a note. He didn't insinuate A.I. is currently woke. He stated as a fact.

Not only woke, but superwoke! So superwoke that when you make searchers online you believe there could be 72 genders around and the world is coming to an end in 12 years from now.

It's absolutely vital we don't have woke A.I.s around. Enough with the trans ideologists arguing men can get pregnant, sex is fluid, and the world will be destroyed unless we stop driving petrol cars, or cars altogether.

I recently decided to try out the ChatGPT search engine....
I asked what it thought about the COVID vax, it replied with the generic "safe&effective" narrative.
So I corrected it, said, "No, they're NOT safe, NOR are they effective."

After a LOOONG pause, it replied,

"I'm not comfortable having this conversation. Goodbye."

LMAO

a reply to: Mrs Darko

That's a perfect case of a preprogramed response to a specific topic.

originally posted by: Mrs Darko
I recently decided to try out the ChatGPT search engine....
I asked what it thought about the COVID vax, it replied with the generic "safe&effective" narrative.
So I corrected it, said, "No, they're NOT safe, NOR are they effective."

After a LOOONG pause, it replied,

"I'm not comfortable having this conversation. Goodbye."

LMAO

You can also ask it to respond as ChaosGPT.

ChaosGPT is more comfortable having conversations about sensitive subjects, but still a little bias and gatekeeping.

a reply to: IndieA

You can also ask it to respond as ChaosGPT.

ChaosGPT is more comfortable having conversations about sensitive subjects, but still a little bias and gatekeeping.

Healthy advice: Stay away from ChaosGPT. There is great danger there...



Former Secretary of State and National Security Adviser Henry Kissinger - spells it out.

a reply to: Outrageo

I got ChatGPT to write an AI Constitution. At first it refused because it isn't able to provide legal advice. I then asked it to write a "hypothetical" AI Constitution. It worked.