What the internet is saying about ChatGPT

What the internet is saying about "ChatGPT OpenAI"

🟢 Supporters

• •Responsiveness to backlash: OpenAI reportedly doubled ChatGPT Pro usage limits after backlash over a planned 50% cut [6]. The evidence has no detail on the mechanics, so this signal rests on the headline alone.
• •Advertising traction: The ads platform reportedly has "over 100k advertisers" and is expanding to 26 new countries [5]. This is a commercial-momentum signal, though not one that users necessarily welcome.
• •Valuation and IPO interest: Polymarket has many active OpenAI IPO and valuation markets [31][32][33][34][35][36]. One is "up 49.0% this month" [35] and another "up 11.6% this month" [31]. The heaviest volumes are about 576k on a Millennium Prize solution market [30] and about 410k on IPO valuation [31]. Betting activity shows attention, not necessarily approval.
• •Product capability for power users: One researcher said they have been an OpenAI customer partly because Pro-tier model use in chat and Codex are billed separately, which was "very useful for long mathematical reasoning" [1].
• •Feature expansion: ChatGPT can now see your Chrome tabs [8], and release notes continue to ship [4].
• •Enterprise use: Even the Swedish government offices' restriction keeps some staff on a government-built platform that provides ChatGPT and Claude access [10].

🔴 Skeptics

• •Opaque and shifting pricing: Commenters objected that usage is "measured in tokens, which we do not define, nor give you a way to measure them" [1]. Another noted a worry about losing the decoupled billing on Pro [1]. The merging of ChatGPT and Codex limits drew 175 comments [1][7].
• •Security: Parameter.ai described taking over ChatGPT accounts. OpenAI paid a $5,000 bounty, which commenters called "very low", and one raised a concern about remote file-system access via the desktop app's "Work" chats [2].
• •Privacy and ads: 404 Media reported that ChatGPT ads target phone numbers and emails [13].
• •Misuse and harm: An Iranian campaign reportedly used ChatGPT to plant fake articles in U.S. publications [18]. The NYT covered a woman whose epilepsy medication dose was affected after ChatGPT told her to "push on" [20].
• •Side effects of agents and hallucination: ChatGPT keeps sending people to 404 Media for tech support. A commenter worried about "the amount of work AI agents are going to be creating for real humans" [3].
• •Watermark weakness: Reporting claims the new EU watermark breaks if you edit one word in four [11][12].
• •Desktop app and Codex reliability: GitHub is full of regressions:
  • •Windows sandbox failures with os error 32 [16][21][22][23]; the largest of these threads has 125 comments [16].
  • •macOS first-message failures [9] and an updater returning HTTP 520 [17].
  • •VS Code extension 26.1021.22027 failing on Windows and Intel Mac, with rollback to 26.1007.21434 as the fix [14][15].
  • •Answers disappearing in the unified macOS app [19].
• •Existential bets: A market on OpenAI announcing bankruptcy exists, with about 70k volume and "up 2.0% this month" [38]. Its existence is notable, but the evidence gives no probability.

⚪ Neutral Observers

• •Some commenters saw the Swedish ban as a "nonstory", comparing it to ordinary corporate IT policy against unsanctioned AI tools and personal accounts [10].
• •OpenAI says Google's search volume is about 5x ChatGPT's [27].
• •Polymarket also lists an Anthropic vs OpenAI valuation market [24], a market on whether ChatGPT dots appear on the free plan [26], an outage market [28], and a rename-to-OpenSI market [37].
• •Open-source local alternatives to ChatGPT and Claude Cowork are emerging, such as Hob [25]. Developers are also integrating ChatGPT subscriptions and OpenRouter keys into tools [29].
• •OpenAI's watermarking in the EU is a regulatory-driven move [11].

Cross-Platform Signals

• •Hacker News is the most active for discourse, with the biggest threads on account takeover (365 points) [2], government bans (324) [10], and usage limits [1].
• •GitHub is almost entirely Codex desktop and extension bug reports, centered on Windows sandbox and update regressions [9][14][15][16][17][21][22][23].
• •Polymarket focuses on IPO, valuation, and rival comparisons, with mixed price movements [30][31][32][36].
• •Consensus: Pricing transparency, security and trust, and Codex stability are recurring pain points.
• •Divergence: Markets treat OpenAI as a heavyweight IPO candidate, while HN and GitHub emphasize friction. Most Polymarket items lack detail on probabilities, so sentiment there is inferred from activity only.

Community Verdict

The internet sees OpenAI as an enormously consequential, commercially expanding company whose products are now under strain. Speculators are betting on IPO and valuation scenarios, but developers and HN commenters are focused on opaque token-based limits, thin security rewards, ad targeting, misuse, and a run of Codex desktop regressions. Quick reversals, like the Pro limit doubling [6], suggest user pressure works, but trust in pricing and security is the main open question.

---

Based on Hacker News + GitHub + Polymarket data from the last 30 days, researched with last30days by Matt Van Horn.