Today’s NYT article about the Amazon “job cuts” phrases it differently than the previously linked gurufocus.com article, as quoted below. Following the news story release, Amazon stock jumped up by 2.5%. In contrast Microsoft, Apple, Google, Meta, the S&P 500, … did not have a notable increase today.
“Executives told Amazon’s board last year that they hoped robotic automation would allow the company to continue to avoid adding to its U.S. work force in the coming years, even though they expect to sell twice as many products by 2033. That would translate to more than 600,000 people whom Amazon didn’t need to hire.”
Stable employee count without hiring employees to match increasing sales seems to be consistent with Amazon’s hiring patterning in recent years. After decades of expansion, employee count seems stable at ~1.5M.
Sounds like the fact that our younger generation isn’t procreating like the folks who produced the boomer generation is actually a good thing, since there won’t be many jobs in the future.
I am more skeptical. With all previous major industrial or technological shifts in USA, I am not aware of any that led to a large decline in available jobs, including looking specifically at entry level jobs. What they do often lead to is a displacement in jobs with some industries seeing a decline, some industries seeing an increase, and some new industries and types of jobs being created that did not previously exist. Compared to the working population as a whole, young persons are often especially adept at adjusting to the changing dynamic, including shifting or transitioning to new or growing employment fields.
We’ve seen this dynamic in recent years with changes in major distribution, including a dramatic increase in majors that are perceived to be related to developing and/or interacting with AI, and a dramatic decrease in humanities majors. We also haven’t seen dramatic changes in overall employment rates so far. As of latest stats I saw from mid 2025, unemployment rates for young persons as a whole are a bit lower than historical averages.
One can certainly find anecdotes about specific companies, such as the earlier stories about Amazon. Amazon may choose to continue to keep employee count stable and not hire 600k persons over next 8 years, or Amazon may lay off 1,500 human resources workers, which is 0.1% of Amazon’s 1.5 million employees. However, Amazon is also hiring in other positions. For example, Amazon recently mentioned they plan to hire 250,000 employees for the coming Christmas rush. Amazon’s website also lists thousands of open long-term positions in other fields, with the largest number of available jobs in software development and operations/support engineering. Skimming through some of these positions, many seem available to newer grads with a tech degree. I am not aware of plans for the total employee count to go down. Instead it sounds like some employment groups within Amazon will decrease in size, and others will increase in size. There are also countless companies that are hiring and will hire in future besides just Amazon, including many companies that do not currently exist.
I see. I was kidding. But u don’t know what is real or not today. Maybe it’s a scam found out and taken down. U can’t trust any job approach like this is my point.
The latter parts of the article are more consistent with my concerns. I had dinner this week with a Sillicon Valley kid who is using AI for all of his coding. He has not lost his job but in fact he might need to hire fewer people to do what his startup needs. I’m surprised that paralegal employment has increased, because law is an area that is really made for LLMs.
One CEO I talk to says no one at his company will lose their jobs because of AI but they will lose their jobs if they don’t use AI. Not clear how enforceable that is but I will be talking about their new AI officer (is this a CAIO)?
I am using ChatGPT, Gemini, Claude and Perplexity in an increasing part of my work. I wrote a review of a proposed journal article on a plane flight today. I read the article quickly and found it a bit frustrating. I asked ChatGPT if the article had justified a couple of its assertions and if so where and how. The article cited someone in 2016 coining an “effect.” I believe I had seen something to that effect described both in one field in the 60s and another field in the 80s. I asked Gemini to find those references (should probably have used Perplexity for that) and it found them.
I am a partner in a venture and one of the other partners is using ChatGPT to lay out the steps for scaling the concept in far more detail than I could ever do. We will be raising capital and those plans make much more concrete how we would do what we say we want to do.
We use AI to turbo-charge Open Source Intelligence. A big improvement over manual work. It either means we can do more for the same engagement or that we can do more engagements with the same personnel.
Our son is an AI developer, architect, and lead. He says the tools have gotten so good on the coding side, neither he nor his team have written an original line of code in a long while. He is also using multiple LLMs to feed problems/solutions to each other. He manages their interaction, watching them “converse," error checking each other as they figure out what to do, how to do it, what to deliver, and then deliver it — at lightening speed. The resulting solutions represent what would formerly take multiple teams of highly skilled developers and consultants months to produce. He says he is currently producing alone what it would take 20 (or more) to do without AI. It’s a sea change in the tech world.
Like you, @shawbrige, DH is using the same combination of AI tools to assist in his consulting work and would agree that they are enabling him to deliver so much more of higher quality in record time. He could easily take on more work but prefers his leisure, and he’s teaching his client to use the tools to eliminate the mundane deliverables not worth paying him to do anymore. Basically, he’s purposefully using AI to eventually erase himself with this client (so he can go fishing).
There definitely is job loss in this new paradigm but, as always, those with the understanding and skills to take advantage of the new technology will continue to prosper. As a society, though, we will need to make sure we can meaningfully retrain/redeploy those displaced as well as teach workers how to integrate these new tools productively into their jobs. Hopefully, the article’s prediction that AI disruption will redraw jobs rather than erase them proves correct, but it is at odds with the absorption rate I posted above.
Oh. Just returned from San Francisco where ShawWife and I delight in taking Waymo. I now check Lyft, Uber and Waymo to compare prices. Each one can be significant less expensive than the others. In one case, we took Uber for $20 when Waymo was $42. In another, Waymo was a few dollars less than Uber or Lyft (and there is no tipping yet on Waymo).
San Francisco is a complicated city to drive in and Waymo drives very conservatively. Waymo knows the city well. Probably safer than a human driver. Driving jobs definitely will go away. But, when I asked ChatGPT and Gemini both told me that autonomous vehicles don’t do well in snow – lane markers will be harder to find, stop signs may be blocked, etc. That may thwart part of my plan for aging – an autonomous car when I am in my 80s and 90s.
https://ride.guru can be a convenient way to get estimated comparisons of Uber, Lyft, taxi, limo, etc.. However, it does not appear to have Waymo in the list of services compared.
@ucbalumnus, thanks. Will try rideguru. I find going from my home to the airport or back, the Lyft/Uber difference can be as much as $40 ($55 versus $95) though last night’s was $3. No obvious reason for the differences. I should factor in my $10/mo statement credit for Lyft (from some credit card) and my 6% credit on Uber from Uber One (from another credit card) – will need to get the app do that for me.
I do have an app that I would like to build using Claude. I may give it a try to see if I can figure out how to do it. Have not done any coding in 35-40 years – but that was how I got spending money and beyond in HS and college and was part of my RA funding in grad school. I might remember enough. I can call some of the kids at startups I’ve coached if I need help (or ask Claude to coach me).
I just check both and if don’t like the price - I wait a few minutes. They change, often rapidly. Pre setting a ride always seems higher than real time but sometimes you have no choice.
The ability to frame the problem/solution well (your consulting skills) is probably all you’ll need; Claude will do the rest (although Gemini may be the better tool for this). You can now build sophisticated apps without having to write a line of original code. This is part of the sea change I noted above and does not bode well for those counting on coding skills to obtain or maintain development jobs. Even full-stack engineers are at risk. The vacuum now is in the deeply experienced AI design/architecting/creative/consulting arena, and most young engineers aren’t coming out of school with skills at this level.
