Toggle navigation
Toggle navigation
Home
DCUM Forums
Nanny Forums
Events
About DCUM
Advertising
Search
Recent Topics
Hottest Topics
FAQs and Guidelines
Privacy Policy
Your current identity is: Anonymous
Login
Preview
Subject:
Forum Index
»
Money and Finances
Reply to "What happens to stocks, investing and generally capitalism when automation and robots do everything?"
Subject:
Emoticons
More smilies
Text Color:
Default
Dark Red
Red
Orange
Brown
Yellow
Green
Olive
Cyan
Blue
Dark Blue
Violet
White
Black
Font:
Very Small
Small
Normal
Big
Giant
Close Marks
[quote=Anonymous][quote=Anonymous][quote=Anonymous][quote=Anonymous]You ever wonder what logically could happen to the entire concept of capitalism? As I save money and buy tech stocks and so forth I am like “will this be worth it in 20 years given ai, robots and automation”? I mean think about how we’ve got “agentic assistants” at Saleforce causing white collar job losses, LLM progress every day, so much ai, and companies are now starting to release robots that can lay bricks and do laundry. What happens to human jobs? In turn, what happens to the concept of owning stocks or investing? How does the advent of humans being second class to more qualified robots work and will companies values still grow? How will people be able to afford stocks if they have no jobs? I asked this of Grok and this was the response: Here’s a clear-eyed, non-utopian take on where this is probably heading over the next 10–30 years. 1. Most current white-collar and many blue-collar jobs disappear or shrink dramatically • Agentic AI (like Salesforce’s Agentforce, Adept, Anthropic’s computer-use models, etc.) is already automating large chunks of knowledge work: support tickets, basic legal research, accounting reconciliation, software development, radiology reads, etc. • Physical robots are crossing the cost-effectiveness threshold fast. Boston Dynamics/Figure/Hyundai, Tesla Optimus, 1X, Agility, Unitree, etc. are all targeting $20–40k price points within 3–8 years with 5–10× human speed on repetitive tasks. Bricklaying (Construction Robotics, Monumental), folding laundry (FoldiMate-style + dexterous hands), elder care, warehouse work, fast-food cooking, trucking (after regulatory approval) — most of these go to zero marginal labor cost once the hardware is paid off. Result: 40–70 % of current jobs (depending on whose study you like: Goldman 2025 says 300 million globally, Oxford 2013 said 47 % in the US, McKinsey keeps revising upward) become either fully automated or “human-in-the-loop for exceptions” that supports 1 human overseeing 20–100 robot equivalents. 2. Company profits explode, stock markets keep going up (for a while) • Labor is 50–70 % of costs in most industries. Replacing it with capital that costs 5–10 % of wages per year causes operating margins to balloon from 10–20 % to 60–90 % in many sectors. • We’ve already seen this preview in software: Microsoft/GitHub Copilot turns one senior dev into 3–5× output → margins go from 35 % to 45 %+. Physical world version will be bigger. • So yes, the S&P 500, Nvidia, TSMC, Tesla, the robot companies, the energy companies feeding the data centers, the copper/uranium/lithium miners — all keep compounding at absurd rates for another decade or two. The people who own capital capture almost everything. 3. The “how do unemployed people buy stocks?” problem This is the core contradiction. You have three main historical ways societies have solved “not enough jobs”: a) Massive political redistribution (UBI, sovereign wealth funds, “robot tax,” negative income tax, state ownership of the robots, etc.) b) Massive make-work / service-sector bloat (the 20th-century solution: everyone becomes a therapist, diversity officer, yoga instructor, OnlyFans creator, etc.) c) Social unrest / collapse / authoritarianism (the bad one) We’ll probably get a mix of (a) and (b) with some places flirting with (c). Current realistic paths people talk about: • Alaska-style permanent fund but national/global: every citizen owns one “share” of the nation’s robot/AI capital, paid as a dividend (Yang’s “Freedom Dividend,” Sam Altman’s Worldcoin idea in richer form, etc.). • Very high wealth / capital-gains / corporate taxes that fund UBI (the Nordic model on steroids). • Government becomes the employer of last resort in “human” jobs (care, arts, local services) that are deliberately kept human and subsidized. • Ownership democratizes: 401(k)s, sovereign funds, and retail brokerage keep growing; people live off a mix of UBI + modest investment returns. 4. Who actually owns the stocks in 2040–2050? Top 1 % already own ~54 % of U.S. stock market wealth (Fed data 2024). That number will probably climb to 70–80 % unless policy intervenes hard. The bottom 50 % own <1 % today. Without redistribution, the median person will have almost no stock exposure and live entirely on whatever transfer system exists. So the realistic outcomes are: Best case (20–30 % probability): coordinated policy response → citizen endowments or UBI funded by 30–50 % tax on capital returns → most people live modestly comfortable lives with a lot of free time, own some index funds through auto-enrollment, and the economy keeps growing. Middle case (50 % probability): patchwork UBI + continued financialization → top 10 % own almost everything, bottom 60–70 % live on $20–40k/year equivalent transfers, gig work, and family money. Markets still go up, but politics gets ugly. Worst case (20–30 % probability): policy gridlock + inequality → markets still go up for owners, but mass unemployment + resentment → authoritarian backlash, capital controls, nationalizations, or worse. Bottom line for you right now If you’re accumulating tech stocks, robot companies, energy, and broad indexes over the next 10–15 years, you’re betting on the “profits explode” part, which looks extremely likely. The political response is the wildcard, but even in fairly aggressive taxation scenarios (say 50 % capital-gains + wealth tax), the pre-tax returns will be so high that post-tax returns still beat almost every historical asset class. So yes — keep saving and buying, but be mentally prepared that “owning the robots” might eventually become partially socialized. The next decade or two of compounding still looks phenomenal for anyone who has capital today. [/quote] Change with AI will be real and certain jobs will require less people but new jobs will be created. None of what you suggest is really likely though. The changes will be big but managable. [/quote] “New jobs will be created”. That’s your take? That’s heavy.[/quot Yes. When cars took over for horses blacksmiths were out of work as were farms that raised and sold those horses. But then mechanics were needed and people at car dealers. Same is true with every major change. This will be no different. [/quote]
Options
Disable HTML in this message
Disable BB Code in this message
Disable smilies in this message
Review message
Search
Recent Topics
Hottest Topics