Thoughts on AI…

I just returned from a technology conference in Las Vegas and had ample time to reflect on the current state of AI. I’d like to share these thoughts as I’m conflicted about recent AI advancements.

Allow me to rewind the clock to the end of 2022 when I first was introduced to OpenAI’s ChatGPT. I was amazed and excited about this technology. While it’s not new, the use cases were narrow and many times limited to research and academic purposes. That brings me to another point.

Transformers that leverage multiple modalities were generally not available to the public and they open a pandoras box on deepfake content. The ability to capture a few seconds of a persons voice and replicate it (AI voice cloning) with precision is impressive to say the least.

However, these new capabilities also expand the risk of using this technology for nefarious purposes.

That panicky call from a relative? It could be a thief using a voice clone, FTC warns

That brings me to some profound thoughts I had while on my flight back from the conference.

1. 150+ years ago Ada Lovelace proclaimed that no matter how powerful “computers” get, they will never be capable of becoming a “thinking” machine.
2. Alan Turing 100 years later sort to define what exactly a “thinking” machine is and this became known as “Lady Lovelace’s Objection”. The “Turing test” or imitation game is a test of a machine’s ability to exhibit intelligent behavior equal or indistinguishable from a human.
3. 75 years later, we are at a point where it’s becoming harder and harder to make this distinction thanks to the recent advancements in AI.
4. Since the 1956 Dartmouth “AI” conference, it has been predicted that a “major breakthrough” was always 20 years away from proving Ada wrong, it’s 2023 and folks embedded into the scientific field of AI as still singing that same tune. We are 20-30 years away from the elusive “AI singularity” event.

Our brain works more like an analog device (yes, no, maybe, probably, and other nuances) vs a digital one (binary one & zero). Perhaps quantum computers will afford the next major breakthrough in AI and getting one step closer to a real “thinking” machine. For now, current AI is great at the difficult things, but terrible at the simple things.

This leaves me with two thoughts to wrap this up.

1. The human brain is amazing and very power efficient
2. Perhaps the partnering of humans and computers is the best possible outcome and not necessary one trying to replace or replicate the other…

“Human ingenuity will never devise any inventions more beautiful , nor more simple, nor more to the purpose than nature does” – Da Vinci