Framing AI debates as a schism between people worried about AI going rogue and those illuminating actual harms is ahistorical and obscures important research
In two recent conversations with very thoughtful journalists, I was asked about the apparent ‘schism’ between those making a lot … Bluesky: @abeba.bsky.social , @mmitchell.bsky.social , and @emilymbender.bsky.social Mastodon: @timnitGebru@dair … , @emilymbender@dair … , and @DrYohanJohn@fediscience.org Twitter: @abebab , @kevinmarks , @mmitchell_ai , and @cher0x801 Bluesky: Abeba / @abeba.bsky.social : This is an excellent piece on the misleading framing of “AI ethics” & “AI safety” as two equally valid research endeavours, while the former is grounded in reality, the latter adds no value — https://medium.com/... Margaret Mitchell / @mmitchell.bsky.social : Great new piece from @emilymbender.bsky.social just dropped! Explains AI ethics and how this is not a parallel to AI Safety (tm). — https://medium.com/... Emily M. Bender / @emilymbender.bsky.social : Talking about a ‘schism’ between ‘AI ethics’ and ‘AI Safety’ is an ahistorical misframing. Here's why: — https://medium.com/... Mastodon: Timnit Gebru / @timnitGebru@dair-community.social : Thank you to @emilymbender for writing this which I will now be pointing to any time someone wants to talk about “AI safety.” — https://medium.com/... Emily M. Bender / @emilymbender@dair-community.social : Recently, I was asked about the apparent ‘schism’ between those making a lot of noise about fears inspired by fantasies of all-powerful ‘AIs’ going rogue and destroying humanity, and those seeking to illuminate and address actual harms being done in the name of ‘AI’ now and the risks that we see following from increased use of this kind of automation. … Yohan John / @DrYohanJohn@fediscience.org : What a wanker. — “Geoffrey Hinton went on a media tour to ‘sound the alarm’ about AI after quitting Google earlier this year. When a CNN journalist asked if he wished he had stood behind previous whistleblowers such as Timnit Gebru when she was forced out of Google (over the Stochastic Parrots paper) … Twitter: Abeba Birhane / @abebab : “'AI Safety' might be attracting a lot of money and capturing the attention of policymakers and billionaires alike, but it brings nothing of value.” @emilymbender nails it! https://medium.com/... @kevinmarks : You making systems smarter than yourselves is like the people who make cryptography they can't crack. By ignoring all the real research in AI and hiring doomers you're guaranteeing that https://medium.com/... https://twitter.com/... @mmitchell_ai : Great new piece from @emilymbender just dropped! Explains AI ethics and how this is not a parallel to AI Safety (tm). https://medium.com/... Cher Scarlett / @cher0x801 : Longtermism philosophy is a distraction from reality. We live in a land of endless reusable resources in the age of abundance, yet there are people starving and sleeping on the concrete downtown. We're in a utopia using a dystopia to maintain it. https://medium.com/... Expand More For Next Unexpand More For Next
Context & Ripple Effects
Emily M. Bender has been building this argument for years: her 2020 paper with the 'stochastic parrot' critique first drew the line between what LLMs do and the intelligence they are marketed as having. The new piece extends that critique from models to the discourse itself, rejecting the 'schism' framing that treats AI ethics and AI safety as two equally valid camps.
The timing matters because the safety framing has already shaped public debate — the pause-training open letter showed how existential-risk rhetoric can dominate headlines while its signatories walk back or turn out to be fabricated. Bender's argument lands amid counter-pressures too: Platformer's critique of complacent AI skepticism pushes the other direction, insisting AI's dangers are real. Endorsements from Abeba Birhane, Timnit Gebru, Margaret Mitchell, and Cher Scarlett signal the ethics camp is closing ranks around a shared framing.
First-order effects
- Bender, Birhane, Gebru, Mitchell, and Scarlett now share a coordinated public line: 'AI Safety' attracts funding and attention while adding no value, a claim that puts journalists' both-sides framing templates under direct attack.
- Researchers and funders associated with AI Safety face a legitimacy challenge from named critics, not anonymous ones — Birhane's echo and Gebru's public endorsement make the critique a camp position rather than one scholar's view.
Second-order effects
- The dispute over the pause letter's credibility gives Bender's side ammunition: if safety rhetoric's flagship intervention was partly fake signatories and walkbacks, the funding-and-attention asymmetry Birhane describes becomes harder to defend as meritocratic.
- Journalists covering AI must choose a frame — existential risk versus present harms — and pieces like Platformer's show the skeptic camp is not monolithic, forcing more granular coverage of who fears what and why.
Third-order effects
- If the ethics camp's framing wins out in policy circles, governance attention shifts from controlling superintelligence toward regulating present-day harms — the same pivot the 'AI as a normal technology' argument urges, suggesting an emerging intellectual coalition between discourse critics and policy pragmatists.
- Whether AI Safety institutionalizes as a funded discipline or gets reframed as a marketing layer over ordinary product risk is now a live contest over research agendas, hiring, and where grant money flows.
The trend: The AI discourse is consolidating into a fight over framing itself, with the ethics camp arguing that existential-risk narratives crowd out research on documented, present-day harms.