US tech stocks see sharp falls over AI cost fears
US tech giants are facing sharp stock declines as nearly $1 trillion in AI spending raises doubts about whether massive data center, chip, and model investments can ever be recouped. Commenters argue that AI is already reshaping user behavior and cannibalizing legacy revenue streams like search ads and social traffic, even as the economics of running large models look far worse than traditional software. Many see a brewing bubble driven by debt-fueled infrastructure buildout and falling PC affordability, with some hoping an early correction will reset unrealistic expectations while others warn of broader financial and social fallout if the AI boom unwinds.
AI Search Quality and Role in Web Search
- Experiences with Google’s AI search are sharply mixed.
- Some report frequent factual errors, outdated info, and dangerous health advice.
- Others find it “directionally right” and a major usability improvement over SEO-cluttered classic results, especially for basic queries and quick calculations.
- Several commenters treat AI answers mainly as triage and reference-finders, not authoritative sources, emphasizing the need to check citations.
- Concerns raised about AI confidently amplifying single Reddit reviews or niche speculation as “people say…”.
Hardware, RAM Shortages, and Data Center Build‑Out
- Intel criticized for under-supplying consumer GPUs (ARC), but others note constraints: use of TSMC, lack of in‑house memory, RAM/GDDR shortages, and questionable profitability.
- Debate on whether Intel should re-enter DRAM or similar memory businesses; most argue it’s a bad commodity bet.
- Broad worry that AI datacenter demand is driving up RAM and GPU prices, depressing PC sales and making new machines less attractive.
- Discussion of “models on a chip” vs. large datacenter models; some see on‑device inference as inevitable, others see ASIC approaches as dead‑ends.
Stock Sell‑Off, Bubble Risk, and Politics
- Many frame the tech drop as an earnings story: huge AI capex, slower-than-hoped ad/search growth, and fears that legacy cash cows no longer backstop AI spending.
- Some view this as a healthy correction in an obvious AI bubble; others call it an overreaction that creates buying opportunities.
- Political uncertainty and elections are mentioned as a factor by some, but others see that as tangential or speculative.
Business Model Disruption: Search, Social, and Content
- Concern that AI summaries reduce click-through and ad revenue for Google, Reddit, and others.
- Roblox user softness attributed more to age-gating friction than AI.
- Worry that if AI consumes content without sending traffic back, incentives to create original content erode.
AI Costs, Pricing, and User Dependence
- Many individual users feel deeply integrated with AI tools and dread price hikes; some say they’d pay substantial amounts, others cap value at ~$20–30/month.
- Observation that “value models” are gaining share, suggesting “good enough” LLMs may dominate as organizations optimize costs.
- Skepticism that current low prices are sustainable; fears of future price increases once dependence is locked in.
Changing User Behavior and Social Media Decline
- Several report spending less time on Reddit, YouTube, and general web browsing, and more time interacting with LLMs.
- Reddit’s login walls and app-push tactics are widely blamed for traffic declines and user exodus, alongside API and UX decisions.
- Some describe using AI as a quasi-therapist or companion, raising questions about loneliness, privacy, and long‑term social effects.