AI 导读
《华尔街日报》提炼了比尔·盖茨一篇 5,784 字 AI 长文的三个要点,核心在于 AI 冲击就业的速度。该文称最大经济风险是速度,AI 可能在新岗位出现、政府拿出认真方案之前就取代大量初级和中级岗位。盖茨希望政府用税收和更强的社会保障网吸收这一冲击,同时认为社会可能需要有意决定哪些类别的工作应保留给人类,并提议设立人类保留类工作、对 AI token 和机器人征税。
正文
3 takeawys by WSJ on Bill Gates's long AI warning.
- The biggest economic risk is speed: AI could remove large numbers of entry-level and mid-level jobs before new ones appear and before governments have any serious plan for it.
- Gates wants governments to use taxes and stronger safety nets to absorb that shock.
- At the same time, he says societies may need to deliberately decide where humans should remain irreplaceable for some categories of jobs.
Bill Gates just warned AI could trigger mass unemployment In a super long (6,000-word) post that @BillGates just published, he says "Even under the best circumstances, the transition to this new AI era will be one of the most turbulent times in human history" Some takeawys - AI is moving too fast for the usual technology-transition playbook. Unlike PCs, AI already runs on devices people own, communicates in natural language, and can learn from the same materials used to train humans, so adoption does not require decades of new hardware, software, and user retraining. - He would probably support a credible global effort to slow AI development, but does not believe geopolitical and economic competition will allow one. - His biggest economic fear is permanent job destruction, not temporary unemployment. Gates argues that AI can substitute directly for human cognition, putting entry- and mid-level work in sales, customer support, software engineering, paralegal work, finance, healthcare and other fields at risk, while increasingly capable robots could start competing for construction and hospitality tasks before 2030. - The decisive point comes when AI becomes reliable enough to work without human checking, because removing the human then becomes economically attractive. - He sees AI as a force multiplier for both attackers and powerful institutions. Lower-skill criminals could gain capabilities for cyberattacks, fraud, deepfakes and potentially biological threats, while governments could use cheaper autonomous weapons, surveillance and opinion manipulation; beyond misuse by humans, - Gates also leaves open the longer-term risk that increasingly capable AI systems could act against human interests. - He worries that always-available, agreeable AI companions could reduce the difficult social interactions through which children build resilience, while heavy reliance on AI for intellectual work could weaken critical thinking; at the same time, he argues that carefully designed AI tutors could improve education by preserving the student's own cognitive effort instead of simply supplying answers. - He also proposes a "Human Reserved" category for work society deliberately keeps human, and taxes on AI tokens and robots to slow some automation.在 X 查看被引用的帖子
来源:@rohanpaul_ai · x.com