LLMs are the excuse businesses are looking for to reduce headcount irrespective of LLM usage/activity/capability. Why employ more people to do an expensive job when headcount and growth are no longer rewarded by investors?
I believe that LLMs provide an excuse for businesses to shed the excess hiring they did during the zero interest rate era. Ultimately, this question comes down to how healthy you think the economy will be in five years, rather than being related to LLMs specifically. I believe there is an AI bubble, so to me this question is asking how much of our economy is mis-invested in AI, and how quickly the rest of the economy can recover if the AI portion implodes. I'm really concerned that the level of mis-investment is high enough that when the bubble bursts it will do severe damage to the larger economy. For example, I'm worried that we'll suddenly have a bunch of abandoned partially constructed data centers, power plants, and fabs. Can the market absorb a massive glut of DRAM and NAND that was contracted for by an entity that suddenly went bankrupt? What will happen to all those warehouses full of GPUs that were waiting for rack space and power?
All of the above could mean we're facing an economic depression that will be talked about for generations. Or it could mean that it will temporarily be unimaginably cheap to start a compute-intensive business, and they will all need to hire developers.
I voted for the middle option because to tell you the truth, I dont think it's possible for anybody to predict where the next 5 years will go. For those who do, polymarket et al exists!
LLMs are the excuse businesses are looking for to reduce headcount irrespective of LLM usage/activity/capability. Why employ more people to do an expensive job when headcount and growth are no longer rewarded by investors?
I believe that LLMs provide an excuse for businesses to shed the excess hiring they did during the zero interest rate era. Ultimately, this question comes down to how healthy you think the economy will be in five years, rather than being related to LLMs specifically. I believe there is an AI bubble, so to me this question is asking how much of our economy is mis-invested in AI, and how quickly the rest of the economy can recover if the AI portion implodes. I'm really concerned that the level of mis-investment is high enough that when the bubble bursts it will do severe damage to the larger economy. For example, I'm worried that we'll suddenly have a bunch of abandoned partially constructed data centers, power plants, and fabs. Can the market absorb a massive glut of DRAM and NAND that was contracted for by an entity that suddenly went bankrupt? What will happen to all those warehouses full of GPUs that were waiting for rack space and power?
All of the above could mean we're facing an economic depression that will be talked about for generations. Or it could mean that it will temporarily be unimaginably cheap to start a compute-intensive business, and they will all need to hire developers.
I voted for the middle option because to tell you the truth, I dont think it's possible for anybody to predict where the next 5 years will go. For those who do, polymarket et al exists!
I think there will be more people working in software than there are today, but they will not be the same people.
People who struggle with the product side will be pushed out.