Good (but scary) overview of state of AI in the og article, some points:
-
The transformation of work by AI is already underway, not a distant future event. The NBER model quantifies this as a potential 366% productivity boost coupled with a 23% employment reduction, with half the job displacement happening within five years. This underscores the urgency of being intentional about how AI reshapes work environments.
-
The future of AI in the workplace favors augmentation over replacement. Most workers prefer AI to automate repetitive tasks and to act as partners or coaches rather than substitutes. Enterprise AI strategies should therefore focus on complementing human skills, freeing teams to concentrate on creative and interpersonal value-driving activities.
-
AI adoption and impact are highly role-specific. While some jobs like logistics management face high automation risks, others like hands-on mechanics remain largely unaffected. Similarly, AI is democratizing expertise by leveling the playing field for freelancers, which challenges traditional premium skill valuations. Effective AI strategies require granular, nuanced understanding of these variations to maximize benefits and mitigate harms.
Ai is not making people more productive:
https://arstechnica.com/ai/2025/07/study-finds-ai-tools-made-open-source-software-developers-19-percent-slower/
Ai is a bubble bigger than the dot com bubble:
https://futurism.com/economist-ai-bubble-worse-dot-com-implosion
https://www.wheresyoured.at/the-haters-gui/
The research about open source developers using ai coding merits a deeper read - it is actually fascinating - they were doing worse, but felt that they have been more productive.