From time to time, I feel compelled to comment on something in the news. The sudden rise in popularity of a new player in the AI arena is such an occasion.
A Purdue University research team reported at last month’s Computer-Human Interaction conference that ChatGPT gives incomplete or wrong answers 52% of the time.
Even more concerning is the fact that China’s DeepSeek by its own admission fares even worse at only 24.9% correct!
Especially alarming as DeepSeek has become the #1 downloaded free app in both Google Play and the Apple App Store. This sent Wall Street in a tailspin as tech stocks fell by more than 10%. What follows are screenshots of DeepSeek’s own rating for various AI benchmarks.
DeepSeek is actually telling us that it is woefully inadequate at providing the public with accurate information 75% of the time! As we become more accepting of this largely unfettered technology, for some inexplicable reason we seem content to accept what we wouldn’t accept otherwise.
Would you honestly continue to use your physician if he were only correct 24%, even 48% of the time?
Recently, Internet Matters found that 25% of our children use AI as a source to research or completely do their academic work. The report goes on to say, ‘Internet Matters in November 2023, estimates that a quarter of children are now using AI for schoolwork, and in total four in ten children are engaging with AI tools — including half of all 13–14-year-olds.”
I don’t believe we have begun to realize the unintended consequences as a result of the rapid development of AI. In our race to be first, we seem to be unconcerned with the chaos in our educational system and other facets of society that this new tool could cause.
Additionally, We recently were warned by the World Economic Forum (WEF) that the world stands to lose 84 million jobs due to AI over the next five years. A figure that is conservative by many.
There are many other ways in which the unrestricted development of AI could have a negative impact on society. Privacy concerns, weapon automation, market volatility, and socioeconomic inequity are among them. Still, AI keeps advancing at breakneck speeds. The bottom line is that there seems to be little consideration for what the future with AI may look like.
I must admit, I first thought AI was harmless. I even went through a period of time where I relied heavily on it. But the more I have learned of the potential threats it might have for society and overall trustworthiness of the end product, I become leary of it. I am unwilling to gamble on my children’s future in order to save a little time now.
I’m certain AI will continue to be developed despite my concerns. We have opened the proverbial Pandora’s box and being unable to control all the players (specifically China,) we no longer have a choice but to press forward. U.S. national security is at stake. The best we can do now is educate ourselves about the potential side effects and prepare for the possible fallout.
Next-generation Generative AI is expected to be with us by 2028. superintelligence AI follows before we elect another president. Superintelligence AI will become capable of independent thought and will be smarter than even the brightest of human minds. It is at this point where there is fear that having achieved autonomy, it could decide that human existence is harmful to the planet.
I will likely stand to benefit short term. Great strides are being achieved in all areas of healthcare, including spinal cord research. However, I am not going to sell my soul though, even if it means gaining the ability to walk and talk once again.
I realize my abstention from the AI revolution will have no effect on the rate at which it will move forward I will nonetheless remain true to my principles. I have become adept at spotting something that is AI-written and I will always be skeptical of something produced by it. In fact, I won’t read anything I suspect produced by it.
