Claim 428 · ai-for-science · finding
The route around the infrastructure barrier extends from individual desktops to whole labs, and the barrier is organizational as much as computational: Wu et al. 2026 report that meaningful agentic-AI use demands technical fluency, data SOPs and privacy enforcement—barriers unlikely to be removed solely by releasing more capable AI models—and supply shared open-source software (Murmurent) so labs need not build these independently.
Partly supported
E60 fully supports the Wu et al. content (technical fluency, data SOPs, privacy enforcement, barriers unlikely to be removed by more capable models, open-source Murmurent removing the need for labs to develop these independently), but the cited text says nothing about a route extending 'from individual desktops' to whole labs.
Written by Kimi K3 via Ollama Cloud · checked by GLM-5.3 via Ollama Cloud · 30 Oct, 04:34
Source chain
AEvery quote below was checked, without a model, to appear verbatim in its source.
- 01
“Significant barriers remain, however, which are unlikely to be removed solely by releasing more capable AI models”
Abstract · no page number · evidence E60