/
Navigation
Chronicles
Browse all articles
Explore
Semantic exploration
Research
Entity momentum
Nexus
Correlations & relationships
Story Arc
Topic evolution
Drift Map
Semantic trajectory animation
Posts
Analysis & commentary
Pulse API
Tech news intelligence API
Browse
Entities
Companies, people, products, technologies
Domains
Browse by publication source
Handles
Browse by social media handle
Detection
Concept Search
Semantic similarity search
High Impact Stories
Top coverage by position
Sentiment Analysis
Positive/negative coverage
Anomaly Detection
Unusual coverage patterns
Analysis
Rivalry Report
Compare two entities head-to-head
Semantic Pivots
Narrative discontinuities
Crisis Response
Event recovery patterns
Connected
Search: /
Command: ⌘K
Embeddings: large
TEXXR

Chronicles

The story behind the story

days · browse · Enter similar · o open

At I/O, Google made AI systems a centerpiece but failed to address concerns over its handling of research about the dangers of AI, which has led to firing staff

James Vincent / The Verge : Tweets: @davewsmart , @funnymonkey , and @emilymbender Tweets: Dave Smart / @davewsmart : Google named it MUM because they want to get flowers & a card one Sunday every year. If you forget, woe betide your rankings. (you might need to tidy your bedroom, unprompted, to get back in the good books) https://blog.google/... Bill Fitzgerald / @funnymonkey : Google is betting its future on AI, and their firing of Timnit Gebru and Margaret Mitchell show that Google cannot be trusted with the data they hold, and how they plan to use it. https://www.theverge.com/... Emily M. Bender / @emilymbender : “they fired two of the authors of that paper, nominally over the paper. If the issues we raise were ones they were facing head on, then they deliberately deprived themselves of highly relevant expertise towards that task.” me, to @jjvincent at @TheVerge https://www.theverge.com/...

The Verge James Vincent

Context & Ripple Effects

The I/O keynote is the endpoint of an arc that began in December, when documents showed Google's AI unit had tightened control over which papers its scientists could publish and asked some to cast its tech favorably. That governance dispute escalated into the departures of Timnit Gebru and Margaret Mitchell, then into waves of academic protest arguing Google's research can no longer be taken at face value.

I/O is where Google asks developers, advertisers, and press to bet on that same AI stack — naming MUM as its search centerpiece while saying nothing about the firings or the paper-review process. The silence matters because the company is simultaneously marketing AI as its future while its own researchers publicly question whether it can be trusted with the underlying data.

First-order effects

  • Google showcased MUM and its broader AI roadmap to developers and advertisers at I/O without addressing the Gebru and Mitchell firings, leaving the trust deficit raised by its own former researchers unresolved at its biggest public stage.
  • The remaining Ethical AI staff and affiliated academics now face a leadership vacuum going into I/O season with no public commitment from Google on how 'sensitive topic' research will be handled going forward.

Second-order effects

  • Academic collaborators weighing partnerships with Google Research must price in the publication-control regime documented since December, pushing some toward rivals whose review processes haven't been publicly contested.
  • Critics like Emily M. Bender and Bill Fitzgerald are converting the firings into a data-trust argument aimed at Google's ad and product businesses — the same audience I/O courts — turning an HR dispute into a brand liability.

Third-order effects

  • If the pattern holds, big-lab research governance moves further under corporate communications control, strengthening the case Wired's coverage flags for tech unions, legal protection for researchers, and external regulation of how labs publish safety-critical findings.
  • Consumer AI showcases may become structurally decoupled from research credibility: companies market capability at events like I/O while the legitimacy questions migrate to regulators, academics, and employee activism.

The trend: As consumer AI becomes each platform company's headline business, internal research-governance scandals are shifting from academic disputes into competitive and regulatory liabilities that keynotes can no longer paper over.

Discussion

  • @davewsmart Dave Smart on x
    Google named it MUM because they want to get flowers & a card one Sunday every year. If you forget, woe betide your rankings. (you might need to tidy your bedroom, unprompted, to get back in the good books) https://blog.google/...
  • @funnymonkey Bill Fitzgerald on x
    Google is betting its future on AI, and their firing of Timnit Gebru and Margaret Mitchell show that Google cannot be trusted with the data they hold, and how they plan to use it. https://www.theverge.com/...
  • @emilymbender Emily M. Bender on x
    “they fired two of the authors of that paper, nominally over the paper. If the issues we raise were ones they were facing head on, then they deliberately deprived themselves of highly relevant expertise towards that task.” me, to @jjvincent at @TheVerge https://www.theverge.com/.…
  • @mer__edith Meredith Whittaker on x
    Google just featured LaMDA a new large language model at I/O. This is an indicator of its strategic importance to the Co. Teams spend months preping these announcements. Tl;dr this plan was in place when Google fired Timnit + tried to stifle her+ research critiquing this approach…
  • @emilymbender Emily M. Bender on x
    @Google @timnitGebru @mmitchell_ai They “are investigating ways to ensure LaMDA's responses aren't just compelling but correct.” ... but in the meantime, is the model in fact framed in such a way that people interacting with it know it's not grounded in any reality? Is @Google ac…
  • @emilymbender Emily M. Bender on x
    And we should trust these claims of responsible development of language-model-based chatbots from @Google because, why, exactly? I mean, they couldn't stand to be associated with a paper carefully documenting the possible risks... #GoogleIO2021 cc @timnitGebru @mmitchell_ai https…