Dots

Commenters debated OpenAI’s new “Dots” as part of a larger scramble around always-on personal agents. Some praised OpenAI’s underlying models and Codex subscription for value and portability (as one commenter reported continuing heavy use), while others accused labs of leveraging early generosity into paywalled, unnecessary products - drawing parallels to Anthropic (johnfahey) and warning that demos often get nerfed by the time features ship (mvkel). Several people worried the product boundaries between Codex, ChatGPT Work, Muse and Dots are blurred (wxw), with Muse seen by some as a stronger consumer play because Meta can subsidize it. A few commenters envisioned these agents ending the PC era by moving compute and users fully to the cloud (jameslk) and suggested open-source alternatives for those worried about vendor lock-in (resiros).

Privacy, marketing and social impacts dominated responses. Many demanded absolute privacy guarantees before trusting always-on agents (teekert, teekert again), and some feared proactive background research could escape sandboxes or create legal exposure (Imnimo). Others found the ads creepy, infantilizing or dystopian (mccoyb, goda90, jason_zig), and worried about vendor lock-in, incompatible agent ecosystems, and potential job displacement (solarkraft, cavoirom, Animats). A few commenters still saw practical value - citing Instinct or recommending Apple as a privacy-minded host - while noting price, availability and name-branding choices will shape adoption (cartersj, jdlyga, tracyhenry).