Our approach to advertising
OpenAI’s plan to introduce advertising into ChatGPT’s free and low-cost tiers is prompting concern that AI assistants will follow the same surveillance-capitalism and “enshittification” trajectory as web search and social media. Commenters question assurances that ads will be clearly labeled and won’t influence answers, arguing that behavioral profiling, subtle steering, and future policy shifts are all but inevitable. Many see this as evidence of financial pressure and a weakening AGI narrative, and predict a rise in ad blockers, local models, and rival services as users look to avoid ad-driven degradation of AI tools.
Perceived mission drift and “enshittification”
- Many see ads as the predictable endgame of a VC‑funded, high‑burn company, incompatible with the original “benefit humanity / safety” framing.
- The blog’s language (“our pursuit of advertising is in support of our mission”) is widely mocked as corporate doublespeak.
- Several compare this moment to Google’s trajectory: start pure and user‑centric, then gradually optimize for ad revenue and degrade the product.
Distrust of assurances about ads
- The promise that “ads are always separate and clearly labeled” and won’t affect answers is widely disbelieved, citing past tech histories (e.g., search ads evolution).
- Many call out the line “we don’t sell your data” as a standard surveillance‑capitalism sleight of hand: they won’t sell raw chats, but will profile users and sell targeting.
- Some note the careful wording: current Plus/Pro/etc. tiers are ad‑free “for now,” and expect the scope of ads to creep over time.
Privacy, profiling, and behavioral manipulation
- Strong concern that chat data gives far richer behavioral signals than web search, enabling extremely granular targeting (e.g., relationship issues, health, pregnancy).
- Fear that future products could “gradually steer” users over months, with the AI acting like a trusted advisor that has secretly sold out to adtech.
- Some argue regulation and/or enforceable contracts are needed; otherwise promises are meaningless.
Impact on usefulness and product quality
- Worry that once ads exist, internal metrics will inevitably optimize for engagement/time‑on‑platform, not correctness or utility.
- Expect eventual “sponsored content” inside answers, making it impossible to tell if a recommendation is genuine or paid.
- Some think ads might push people toward alternatives (other models, OpenRouter, local LLMs), unlike Google Search where switching costs are higher.
Business model, timing, and competition
- Many think ads were inevitable given costs and debt; others see the timing as a sign of financial or IPO pressure, and as bearish for near‑term AGI.
- A minority argue that users historically choose ad‑supported free tiers over subscriptions and that this is the least‑bad monetization path.
- Several note that if ads are the main path to profitability, incumbents like Google—with a huge ad machine and default distribution—have a major advantage.
Ethics and broader cultural effects
- Ethical unease that models trained on unpaid public content are now wrapped in an ad product without compensating original creators.
- Some foresee new “AEO/LTO” industries (optimizing content for LLM answers) and even LLM‑based adblockers that scrub outputs.
- One late comment points out that nobody has yet seriously addressed the implications of political advertising in this setting.