Why Doesn't the US Fund National Open-Source AI Models?
A Reddit user questioned why the US government does not fund and train national open-source AI models, pointing out that recent advancements like DeepSeek show training can cost relatively little while providing massive public benefits. This discussion highlights growing policy debates around state-sponsored AI infrastructure, contrasting private monopoly pricing and foreign competition with public-interest open-source alternatives. The poster noted that models like DeepSeek-V3 were trained for a relatively low cost of around $6 million, arguing that a similar government-backed initiative could solve security concerns and prevent corporate monopolies on token pricing.