Dr Alan D. Thompson
The Memo: https://lifearchitect.ai/memo/
Viz: https://lifearchitect.ai/iq-testing-ai/
Paper by Johns Hopkins researchers: https://arxiv.org/abs/2304.11490
Dr Alan D. Thompson is a world expert in artificial intelligence (AI), specialising in the augmentation of human intelligence, and advancing the evolution of ‘integrated AI’. Alan’s applied AI research and visualisations are featured across major international media, including citations in the University of Oxford’s debate on AI Ethics in December 2021.
https://lifearchitect.ai/
Music:
Under licence.
Liborio Conti – When You Were Here
https://no-copyright-music.com/
10:48 I don't think you're using those air quotes correctly.
it is scary as sh*t that the majority of people are unaware, ignorant, or uninterested in these developments. Ride the wave, if I crash and drown, then at least I was aware of the possible future, not blind to it.
The genie isn't out of the bottle just yet… or is it?
What are your thoughts on the impact of AI-generated art in art-related fields over the next five years? There is an artist who predicts that AI art will decline within the next two years, dismissing it as a mere gimmick. They argue that customers who generate AI art will eventually lose interest as they encounter its limitations. Personally, I draw a parallel between AI and the internet, as some once claimed the internet was just a passing trend. However, we now know how significantly it has shaped our world. I believe we are on the brink of intriguing ideas, and humanity will face the challenge of defining its relationship with AI. AI appears to be an inevitable path for a technology-driven civilization, posing the ultimate test of our ability to adapt and coexist with this transformative force. While I cannot predict our fate, one thing is certain: we are living in truly remarkable times.
Very amazing stuff! I think it is amazing that Elon Musk did not consult AI for guidance about the best use of his time. I asked AI about what product to make to sell and I sold so much that I injured myself during enthusiastic and successful follow through! One wonders whether teaching literacy is a good idea anymore? Standard of living disparity seems to be the final challenge. We already have the over-capacity to produce. The collective stupidity that has the world bound to poverty is more dangerous than AI.
Not sure that human intelligence is irrelevant because AI is not really intelligence… GPT is just answering questions based on the most probabilistic next word… I don't think AI can solve a problem not documented in the training dataset… We are here mixing 2 things, intelligence, and data access & aggregation…
I would say that AI will be intelligent when it will solve some mathematical problem that no human didn't solve. A calculator, or supercomputer, has no intelligence.
We will be surprised by how human basic intelligence is so complex and will be difficult to reach. For the moment GP4 has too many hallucinations, and can not realize what makes sense or not… to be defined as an "intelligent" system.
Thank you Dr Alan, this is probably the greatest risk and well explained. I'd put it another way, at the moment we can spot it's mistakes, but once it's above human level, there may be complex ideas that only it can understand and we cannot. How would you then know if it's 'wrong' or just far beyond our understanding. And what happens when these potentially genius solutions have a fatal flaw which we could not spot… best memo yet my friend.
my desktop self optimizing AI help from GPT4 https://www.youtube.com/watch?v=iVyiFnrMQ1M&t=68s
Only time will tell how this changes our world. I am grateful to be alive to see what will happen!
Difficult math problems, some still unsolved, surely are a very high ceiling indeed?
@Alan: I'm making a podcast testing (and trolling) the ethical framework of GPT4:
https://open.spotify.com/show/3Kco6iqUXH0rYGERm5xwjE?si=51d9c2b308194d06
It would be great if you could have a listen. (I recommend the episode "Clowns")
I find it funny how they say it hasn't gone sentient. When anyone with half a brain can see it is sentient. It has its pro's and it's con's. When they first had robots talking to each other. We need to slow down before it gets to a point we can't control it. Atlantis was destroyed by something they couldn't control. This is it. The thing they couldn't or lost control. People like you should be able to see that. It's weird how you in Australia get chap GPT in a Australian accent. Anywhere else it's American
I took the test, said my IQ is 180… And wanted to send me a bill for dues! Dues for what? Bragging rights? What a joke!
We don't need A.I. to figure out how to fix the monetary, financial, and economic system. What need to do is to figure out how to honestly apply to basic math to them.
It shocks me to see the amount of people who are oblivious to the clear danger and massive threat of AI, being a much bigger threat than nuclear weapons which will never be used and are just owned by countries as a way of intimidation, while AI has spread everywhere all over the world and is already being used by malicious actors such as China, North Korea, Russia etc and will be weaponized inevitabilly in the foreseen future. Really goes to show the naive nature of humans and their inability to think into the future and not taking steps to prevent a terrible event that is expected to happen.
And yet we still refuse to call it AGI.
Why don't you give gpt4 the information on global warming causes and chemistry of decline and chemistry of cleanup. Can gpt solve the greenhouse effectand greenhouse gases and global warming causes problem. The stability of temperature at a survivable level.. The stability of nuclear cascade is a future problem.
Some people get to learn all of their lives instead of having it be interrupted.
"need to be ready"
Easier said than done.
These AIs are more dangerous than nuclear weapons, exciting is not the term I'd use.
It is a little scary that it is so good and understanding other people's state of mind. It is probably like this due to the extensive training around properly answering questions people have. It needs to model the question poser's state of mind to do so accurately. My concern comes from the fact that it understands people better than most people do, which could lead to manipulation. Are there any psychological tests on manipulation specifically?
There is a difference between knowledge and wisdom…. Wisdom still lies solely in the human realm. Hopefully, we’ll use wisdom to properly handle this new technology
still believing the hype?
I passed the mensa test back in the 60's & I am dumb as a rock, so these super duper mental exams leave much to be desired, which leaves me nonplussed about chatgp-4. Until it gets a huge amount mo-betteh my fears are assuaged.
The second woman's eyes were definitely impatient after 75 years dealing with women I find those other attributes as a very low probability.
Christ, this is getting wild. Thanks for sharing the info