11/03/2024
You should save this post, because a majority of your friends will likely find it difficult to believe.
There are four (4) pictures below. The picture on the far left is of a plane. The test subjects were told to look at the plane and then, later, recall the picture of the plane.
The test subjects were hooked up to an EEG (Electro Encephalo Gram) machine, which records electrical activity in the brain and displays it as "waves" on a screen.
Now, enter Artifical Intelligence (AI). Using this data, AI has been able to "read" the waves and essentially reverse-engineer and extrapolate images from the waves. Three examples of the AI results are to the right of the actual plane.
I agree they're not great, but this technique is in its infancy. As developments are happening more and more rapidly, it is without question that this system to read and graphically reconstruct images based on brain waves will improve at lightning speed.
Additionally, the same EEG machine can be connected to read brain activity when one us sleeping. Yes, that means that the AI will be able to graphically reconstruct static and, eventually, dynamic images and sounds from our dreams. AI will also be able to inform us what coma patients are thinking, and a host of other applications.
I'm happy to have been involved to varying degrees in AI since 1995, with my invention of a CertiBot for one fo my companies (SafeCertified.com). If you look at the screenshot from more than 20 years ago, you'll see our gateway homepages and "Kay," the first interactive CertiBot.
We've come a great distance from the days of binary response trees to actual fuzzy logic and generative AI chatbots (like ChatGPT.) Whereas the CertiBot used pre-assembled scripts and replies, and could only integrate new questions, answers and responses based on our data entry, nowadays the large language model (LLM) systems enable the AI to independently integrate and generate human-like conversational responses it may learn from interacting with people.
I will say that the only advantage our CertiBot had over AI like ChatGPT was we would correlate each script (query, response, reaction, etc.) with a human facial status, gesture or position. It wasn't very fluid, but it did provide a more human-pleasing experience. People are, by definition, visually oriented, and being able to "look at someone" allows the brain to better accept and work with the AI. This ingrained habit and preference for visualization is programmed early, when we're babies, and we seek to identify and see our mothers. It's also why we have a habit of "seeing" faces in the sky, or in clouds, or in abstract patterns, that are not actually intentional.
We have an exciting and (we think) important announcement regarding our development in AI that will be disclosed in about 90 days. If you're interested in being put on our distribution list, let me know.
