Safety
The case for boycotting Generative AI: https://garymarcus.substack.com/p/the-case-for-boycotting-generative
Gary Marcus argues that current generative‑AI models fail to meet basic standards of reliability because they routinely hallucinate facts, amplify existing biases, and provide no transparent audit tra
Gary Marcus argues that current generative‑AI models fail to meet basic standards of reliability because they routinely hallucinate facts, amplify existing biases, and provide no transparent audit trail of training data or decision logic.
He contends that these deficiencies—combined with potential economic disruption, privacy abuses, and a lack of regulatory oversight—justify an industry‑wide boycott until robust verification methods, accountability mechanisms, and ethical guidelines are established.
The piece urges developers to prioritize safety and transparency over commercial deployment and calls for regulatory frameworks that enforce data provenance, model interpretability, and user protection.
Related
- @GaryMarcus Marcus isn't wrong. Everyone bolts tools onto transformers because they hit walls fast. Neurosymbolic AI is rising for a reason:…
- update with some context and additional thoughts: https://open.substack.com/pub/garymarcus/p/the-generative-ai-fizzle?utm_source=app-post-st…
- @GaryMarcus, great interview and explanations of the challenges in AI. https://youtu.be/_pSivPlRx5o?is=dlFN6YBXsMA-tW4h
- Receipts: https://open.substack.com/pub/garymarcus/p/deconstructing-geoffrey-hintons-weakest?r=8tdk6&utm_medium=ios
- https://www.irozhlas.cz/zivotni-styl/spolecnost/marcus-jazykove-modely-ve-skutecnosti-inteligentni-nejsou-od-lidskeho-mozku-se_2606201209_ak…
Source: Gary Marcus (X) | 2026-09-09