John Phipps: The Truth About Artificial Intelligence
USFR - CUSTOMER SUPPORT 6/17/23
Frequent correspondent Jay Brown from Ravenna, OH is skeptical of Artificial Intelligence (AI):
“I could take issue with every point you made but instead I will give you the only truth. There is no artificial intelligence now or in the foreseeable future. Mashing up Siri search and databases is not AI. Just sophisticated programs doing what they are programmed to do, nothing more.”
Many people share your dismissal, but often it is centered on the name of the technology itself. The term artificial intelligence was first used in the early 1950’s and rapidly became a standard expression for science fiction writers. Even in the beginning few agreed on exactly what it meant.
One big issue is the confusion with sentience – self-awareness. That is a whole ‘nother scientific and philosophical issue. Today, the AI industry avoids that implication, but the image of the Star Trek character, Data, pops into minds . Worse still, the suggested alternative terms like “machine learning” and “LLM’s – Large Language Models” don’t clarify matters much, so the concept of AI is pretty much up to users.
As far as intelligence, there is a widely used standard – the Turing Test, first proposed by the mathematical genius of WWII, Alan Turing. The test is simple: if during a conversation you cannot identify a human from machine responses, you are dealing with intelligence. This criterion would appear to be approached and often met by current generation chatbots, such as ChatGPT4.
When AI programs began beating the world’s best Go players, the common comment from the masters was the victories were due to moves never imagined in the centuries the game has been played. Those moves were not programmed, only the goal (winning) and rules were.
To the extent that AI is simply a mashup of a search and logic software with extensive memory, how greatly different that is from a basic description of human brains? We use logic to reach conclusions and formulate actions about things we know or experience.
AI is already close enough to human capability that creative occupations, from copywriters to commercial artists to programmers are being laid off in thousands, and computer engineering graduates are stunned by a lack of entry level jobs. My judgments are also based on the same logic as EV’s – that’s where the money is going.
Perhaps some individuals do know the only truth as Jay puts it, but owners of hundreds of billions of dollars clearly embrace other conclusions.