There’s a lot of downplaying AI, and I agree it isn’t there yet. LLMs are already pretty good at spitting out well documented, quality code, and they are rapidly improving. There are already decent “AI” (in quotes because we don’t really have AI yet) tools for evaluating and improving network security. These areas are ahead of others for a couple of reasons. Those developing AI are developers. They are using what they know to develop and test AI: coding. Also, LLMs can be trained in any language, including code.
There are many other fields in which LLMs and eventual AI are a good fit for replacing at least repetitive and mundane tasks.
AI doesn’t need to perform a tremendous amount of labor to displace a large portion of the workforce in a given field. The 80/20 rule is fairly accurate. If AI is able to perform 20% of the work, and it takes a fraction of the time to validate that work as it does to perform it, AI could displace 50% or more of a workforce in many fields.
I think there is a lot of head burying because people don’t want to face the reality that is coming, which is there will be far fewer well paying professional jobs in the not too distant future.
All the questions about how businesses will make money when the workforce drastically shrinks are nonsensical. The top 20% are already driving the vast majority of economic development, which is building of data centers and doesn’t benefit anyone outside of those invested in AI. Yes, many businesses will fail and there will be consolidation. The haves will prosper and the have nots will be made poor and irrelevant. It would be amazing if we could legislate AI to ensure everyone benefits, but if you’re paying attention, you know that won’t happen.
1 Comment
There’s a lot of downplaying AI, and I agree it isn’t there yet. LLMs are already pretty good at spitting out well documented, quality code, and they are rapidly improving. There are already decent “AI” (in quotes because we don’t really have AI yet) tools for evaluating and improving network security. These areas are ahead of others for a couple of reasons. Those developing AI are developers. They are using what they know to develop and test AI: coding. Also, LLMs can be trained in any language, including code.
There are many other fields in which LLMs and eventual AI are a good fit for replacing at least repetitive and mundane tasks.
AI doesn’t need to perform a tremendous amount of labor to displace a large portion of the workforce in a given field. The 80/20 rule is fairly accurate. If AI is able to perform 20% of the work, and it takes a fraction of the time to validate that work as it does to perform it, AI could displace 50% or more of a workforce in many fields.
I think there is a lot of head burying because people don’t want to face the reality that is coming, which is there will be far fewer well paying professional jobs in the not too distant future.
All the questions about how businesses will make money when the workforce drastically shrinks are nonsensical. The top 20% are already driving the vast majority of economic development, which is building of data centers and doesn’t benefit anyone outside of those invested in AI. Yes, many businesses will fail and there will be consolidation. The haves will prosper and the have nots will be made poor and irrelevant. It would be amazing if we could legislate AI to ensure everyone benefits, but if you’re paying attention, you know that won’t happen.