You can click on the copy icon in each video and paste in your web browser's address bar to go straight to YouTube..
This starts off as a great overview of AI. Could play the first 6 of 5 stages.
Stage 1: Rules-Based AI
Stage 2: Context-Based AI
Stage 3: Narrow-Domain AI
Stage 4: Reasoning AI ~ This one is relatively new. Chat-GPT is an example.
Stage 5: Artificial General Intellegence ~ Not sure if this is cutting edge or just a concept still.
Stage 6: Super Intellegent AI ~ Only a theorietical idea.
Stage 7: Self-Aware AI ~ The stuff of movies. It has been speculated that quantum computing may help create this.
I honestly think that it is sheer speculation that this will ever happen. What is more liely to make AI dangerous
is it adding even more system design complexity. We still don't understand consciousness.
There are already airplane accidents that cannot be explained due to the systems involved in the crash being too complicated to troubleshoot.
My first desktop processor had errata (unforeseen side effects of design, not manufacture defects).
Stage 8: Transcendent AI ~ This sounds more like an application of AI instead of a capability classification of AI.
Stage 9: Cosmic AY ~ Just science fiction style speculation. Has more to do with going beyond the Kardashev Scale than describing AI.
Stage 10: God-Like AI ~ Same as Sage 9, but further up the extended Kardashev Scale.
Computers have for decades process faster and multi-task better than us. The limitation is the kind of thinking/processing they can do,
so if they can reach Stage 7, then they could rapidly proceed to Stage 9 (and Stage 10 if physics allows it).
Hit your web browser's back button to comback to this page!
A video that reasonably quickly goes through all seven theoretical levels of the Kardashev Scale.
This is a playlist. Click on the icon inside the upper right corner of the video to see list of videos. Or click on the track buttons at the bottum to jump through videos. Note that only Types 0 through 3 are actualy part of the original Kardashev Scale.
Could play the first half of the video. Does a great job of a explaining aspects of AI in a way that most people can understand.
Could play some of the algorithms. A great run through various machine learning algorithms with simplified explanations. An algorithm is simply a planned sequence of steps to solve a problem. You probably already using machine learning to filter your email. Watching this video brings back memories of me trying to research to make a smart thermostat as my graduation project.
A video about how AI development matches up with how the human brain works and develops abstractions; How AI evolution is similar to natural evolution; evolution went from tinkering with organisms over generations, to organisms learning from experiences within their lifetimes, to using language to learning from the experiences of others. The video uses plain language and has decent illustrations.
Quantum computers work on quantum probabilities of qbits instead of the definitive, discrete binary states of bits.
Quantum computers use quantum probabilities to do many calculations simultaniusly instead of one at a time like binary computers.
This will allow much of current security measures to be broken when such computers work practically.
Quantum computers have huge cooling systems that dwarfs the actual computer hardware far more than my computer,
because they need to cool close to -459°F (absolute zero)
while I only need to stay below 158°F which is still somewhat low for a high-end desktop computer.
Quantum cryptography is sending data that is encrypted with quantum keys that are sent as individual photons.
The keys will not work if they have been observed before they get to their destination.
This video shows quantum cryptography being sent between places via fiber-optic cable and satellites.
A key detail of quantum physics is that qbits exist in probability waves untill you observe them, then they collapse to precise states/values.
The video does not mention this, since it is focused on dedicated communication lines,
but my educated guess is that the satelites must be re-encrypting the data,
and that this will not work on the Internet unless you have trusted, dedicated routers that act like the satellites mentioned in this video.
This would be a big shift, since we generally operate the Internet as untrusted and unsecure.
Your encrypted data flows through the Internet in its encrypted form,
but the basic information about the destinations has to be left unencrypted for routing.
Google is developing quantum computer chips. The disappointing aspect of this video is that the chip did perform an amazing fast calculation, but that calculation was just a benchmark thet has no practical use. It is speculated by some people that the quantum processing capacity is made possible by parallel universes. This sounds like a science fiction story that my collage roommate told me about a computer that grew into extra dimensions to gain computational power and then restarted the universe after the universe faded out and the humanity died out.
How an AI has progressed to solve the problem of predicting protein shapes which is a complex biochemistry problem. My collage roommate told me (back in the late 1990's) that proteins were impossible to predict. Now AI can make successful predictions, but before that, Folding@Home has been making exhaustive numerical simulations that my computer participates in.
A longer and more detailed version of the above video. Has many nice visual explanations.
This video is created by the company that makes all the graphic cards that I bought in the last twenty years, except for my Macbooks. Videos shows the making of a beautiful visual of the COVID-19 Virus. It took people donating the use of thousands of video cards to help work on COVID-19 by running Folding@Home. This is what make Nvidia a big name in artificial intelligence. An exaflop is a million million million variable decimal point operations per second.
The human brain uses as much power as a small light bulb, but does way more computation than computers that use way more energy, so a company is developing a chip that uses neurons.
I did a short paper in school about neural networks back in the 1990's. To the best of my knowledge, this is a good explanation without getting overly complicated.