From a Brave New World to Artificial Intelligence: Are We Living in the Future We Feared?

As someone who is deep in this space, we are living in a world of AI over-hype. AI peaked on Gartner’s AI hype cycle around September of 2023. It is now entering into what is termed the Trough of Disillusionment. AI has promised but not really delivered. It hallucinates, it is not precise, it has significant challenges in completing many tasks. We are entering what is now termed an AI Winter. There is no new AI frameworks, most everyone is developing off of the OpenAI model. Not much innovation. OpenAI has ran out of data to consume as it has consumed the entire internet, and now it is turning to synthetic data using large swaths of people in places like Kenya. Reports are growing that question the value in the near term around AI. Microsoft is starting to receive pressure around their CoPilot AI technology. Varying degrees of value are being seen, with most saying it is not meeting expectations and that the cost doesn’t justify the value. Now you could argue that given enough time, it will be overcome. And that may be true. But we have a precursor to look at already to see what that looks like and it is Self-Driving Cars. It rolled out with great fanfare almost 10 years ago. Over the time it has gotten better, but it has not quite reached the state of general use. It is still plagued with its lack of accuracy and understanding of situations. There is no doubt it has gotten better, but the last 10% is a very large hill to climb.

We will see what the future holds, but today, AI delivers little promise and but shouts a lot of hype. That may be due to the fact that 100’s of billions of dollars have been poured into and hype is what is needed to recoup that money.

Dgsweda wrote:

AI peaked on Gartner’s AI hype cycle around September of 2023. It is now entering into what is termed the Trough of Disillusionment.

Thank you for stating what I had been suspecting. It has not proven all that it was cracked up to be — at least not yet. Michio Kaku, a few years ago, wrote that Androids that really functioned like people are perhaps generations away.

This is not to say that things have not advanced in the AI direction, they have. But there are limits for sure.

Good to have someone on top of things to make these comments! Thank you, DS.

....that with the proliferation of AI-generated documents in academia, the publishers of the old "blue books" have a lot to gain from AI. Smart professors and teachers will jettison online methods for the old methods because the consequences of not doing so are to allow the cheaters to run the roost.

And I tend to agree with David on most of this, as what AI does as a rule so far is to average out what is being said in public. We might joke that it is not AI, but AR, artificial reguritation/repetition, and as such, it lacks, so far, the spark of creativity which is true intelligence.

Maybe someday that will change, but going back to the original situation, one of the easiest ways to recognize AI is that its conclusions tend to be trite, or too close (e.g. Google Gemini) to the biases of those running it. I actually learned to spot cheating due to AI math because the algorithms don't list a lot of the necessary steps, and you'll also get students who make mistakes in the middle of a problem, but after a miracle occurs, get the right answer.