Versions :<12345678Live>
Snapshot 5:Thu, Jun 12, 2025 3:45:24 PM GMT last edited by Nick

Meta Invests $15B in Scale AI for Superintelligence Lab

Meta Invests $15B in Scale AI for Superintelligence Lab

Above: Double exposure photograph of a portrait of Mark Zuckerberg and the Meta Group logo at Kerlouan in Brittany, France on April 11, 2025. Image copyright: Vincent Feuray/AFP/Getty Images

The Spin

Zuckerberg is bringing Meta to the forefront of the AGI push with this massive-scale Scale AI investment and a superintelligence team, positioning his company to surpass OpenAI and Google. By personally recruiting top talent and leveraging Scale's data expertise, Meta will be able to integrate advanced AI into its products, setting a bold path toward AI-human parity.

MetaAGI's partnershipcreation, withwhether Scaleby AIgovernments or companies, arisks companyan accusedexistential ofcrisis. exploitingOutstripping low-wagehuman internationalcognition, workersAGI forcould dataerode labelinghuman identity, agency, and maintainingrelationships, Pentagonleaving weaponsus contracts,dependent raiseson serioussmarter concernsAI assistants. TheUnpredictable dealand alsopotentially appears desperateuncontrollable, especiallyAGI givenmight Meta'sprioritize historymisaligned ofgoals, failedleading venturesto likecatastrophic thesocial metaverse and recentmilitary AI setbacksoutcomes. TheThese focusethical onperils "superintelligence"far seemsoutweigh morethe likepotential expensive marketing than genuine innovationbenefits.

AGIMeta's creation,partnership whetherwith byScale governments or companiesAI, risksa ancompany existentialaccused crisis.of Outstrippingexploiting humanlow-wage cognition,international itworkers couldfor erodedata ourlabeling identity, agency, and relationships,maintaining leavingPentagon usweapons dependentcontracts, onraises smarterserious AI assistantsconcerns. UnpredictableThe anddeal potentiallyalso uncontrollableappears desperate, AGIespecially mightgiven prioritizeMeta's misalignedhistory goals,of leadingfailed toventures catastrophiclike socialthe metaverse and militaryrecent outcomesAI setbacks. TheseThe ethicalfocus perilson far"superintelligence" outweighseems potentialmore benefitslike expensive marketing than genuine innovation.

Metaculus Prediction

There is a 50% chance that the first weakly general AI system will be devised, tested, and publicly announced by Jan. 10, 2027, according to the Metaculus prediction community.


The Controversies



Go Deeper


Articles on this story

Sign Up for Our Free Newsletters
Sign Up for Our Free Newsletters

Sign Up!
Sign Up Now!