Quote · The Prof G Pod with Scott Galloway
Why AI Stocks Are Overvalued + Are Networking Events Actually Worth It?
Where this was said
Q1: Is AI Fatigue Real and Are AI Stocks Overvalued?
At 7:30 · chapter starts 2:07
Pablo's question cuts to the heart of the most pressing tension in tech: companies are spending furiously on AI, but the productivity payoff is stubbornly elusive. Galloway opens with the macro: worldwide AI spending is forecast to hit $2.6 trillion in 2026, up 47% from 2025. But then the counterweight arrives — a 2024 MIT study found humans are cheaper than AI for 77% of vision-based tasks, and a survey of 6,000 senior executives found 90% of firms reporting zero productivity impact over three years [1] — Scott Galloway "No AI productivity impact: A survey of 6,000 senior business executives found that 90% of firms report AI having no impact on productivity …" 04:25 . Even the executives who use AI average just 1.5 hours a week. The picture that emerges is of a technology that works beautifully in demos and lab conditions but struggles to move the needle at organizational scale. Galloway draws a distinction between AI technology, which he's bullish on, and AI stocks, which he believes are priced for a world that doesn't yet exist. He predicts CFOs will soon take the wheel from CEOs who've been using AI talk to wallpaper over poor management decisions, and that the era of experimentation will give way to a hard ROI reckoning [2] — Scott Galloway "We're leaving the AI experimentation era and entering the ROI era — and that transition is going to hurt AI stock valuations hard. Galloway…" 05:40 . The one silver lining: consultants who help companies decide where to spend on AI are about to have a very good few years.
Worldwide AI spending is forecast to total $2.6 trillion in 2026, a 47% increase year over year.
Total AI spending in 2025 was $1.76 trillion, the baseline from which 2026 forecasts a 47% jump.
A 2024 MIT study found that for 77% of vision-based tasks, it is cheaper to employ humans than AI. NVIDIA's own VP of Applied Deep Learning admits that for his team, AI currently costs more than the employees it's supposed to replace. The economics just aren't there yet.
A 2024 MIT study found that for 77% of vision-based tasks, it is cheaper to employ humans than AI.
A survey of 6,000 senior executives found 90% of firms report zero AI productivity impact over three years. Meanwhile AI tools boost individual coder output by 290% on single files but only 30% at the company level — showing that lab results and real-world ROI are two very different things.
A survey of 6,000 senior business executives found that 90% of firms report AI having no impact on productivity over the last 3 years.
69% of firms actively use AI, with higher usage rates at younger, more productive firms, but adoption remains concentrated among small groups.
