Back to Writing Some Observations on Artificial Intelligence

Some Observations on Artificial Intelligence

Apr 11th, 2023

There is a rarely a day when you don't see a headline regarding AI and why you should be afraid of it. I am far more skeptical. That being said I think its worth having some review on the state that AI is in, its trajectory and what elements may be cause for concern.

Where are we?

Across the world people are discussing the next batch of jobs being replaced by AI systems and how everyone should be fearful of this and that. However, all of these LLM (Large Language Models) are nothing more than fancy spell checks or next character guessing engines. Yes, they are getting more complex, and yes they are getting good at observing what "should" come next in a conversation. However, they are still "dumb" in as much as, they are not "thinking" about what fits best or reassessing what they think is right. When they spew out incorrect facts and we correct them, they are NOT fixing the mistake, they are merely "changing the conversation".

Where is this thing going?

LLM systems are a big leap in the growth of AI systems. Yes, there will be research jobs or pencil pushing jobs that are more systematic that may be at risk, but in all likeliness these jobs would merely change rather than be completely discarded. There is no evidence that AI systems are getting definitively "smarter" or gaining any degree of independence or self understanding. They are all still nothing more than basic machines receiving and input and producing an output.

Cause for concern

Is there anything on a larger level that people should be concerned about? In my experience the big winners and losers in the AI world will not be the consumers. Businesses and organizations will be the biggest winners and losers in this domain. Ultimately companies are using AI already to increase productivity or observe hidden trends in their data, companies who are not doing this are losing. Organizations who are using generative AI to get eyes on social media are ensuring they stay in front of people. The organizations that are not doing this are being ignored.

AI systems as headlines get much more attention when they're perceived as being a threat to our livelihoods. However, I have not been convinced that they will any more of a threat to us than the drill was to the carpenter, or AutoCAD was to the architect.