The AI Investment Paradox — A 1962 Book Explains Why Billions Don't (Yet) Deliver
The Paradox
$37 billion invested in generative AI in 2025 alone. A 3.2x increase from the year before [1]. And yet — 95% of businesses have yet to see measurable ROI from their AI investments [2]. 42% of companies abandoned their generative AI initiatives entirely [3].
How is this possible? How can an industry attract this much capital while delivering this little return? I’ve been tracking this disconnect for a while — the “disappointment gap” between AI hype and actual outcomes is real, and it keeps widening.
“By design, the innovation funnel leads to survival of the safest ideas.”
“By design, the innovation funnel leads to survival of the safest ideas.”
Yes been there and seen that. Not only in Germany where we are unfortunate very famous for that 😉.
Two thoughts on that:
Spending more time in ideation and evaluation can make a huge different. Mechanisms like Amazon’s working backwards can bring a lot of understanding already in those early phases of the funnel. Managing risks. Tom, +1 on the power of diverse teams!
✨ It has never been a better time to be excited about the future.
✨ “It has never been a better time to be excited about the future.”
🔍 I missed this interview back in October last year when Jeff Bezos compared today’s AI boom to the internet bubble of the 2000s at Italian Tech Week 2025 [1]. He warned of hype but insisted AI is “real” and will transform every industry. In the interview, he explained why industrial bubbles can benefit society and predicted that AI will raise both productivity and quality worldwide.
🤔 '𝗪𝗿𝗶𝘁𝗲 𝗙𝗶𝗿𝘀𝘁 𝗼𝗿 𝗕𝘂𝗶𝗹𝗱 𝗙𝗶𝗿𝘀𝘁? 𝗪𝗵𝘆 𝗔𝗜 𝗶𝘀 𝗥𝗲𝘄𝗿𝗶𝘁𝗶𝗻𝗴 𝘁𝗵𝗲 𝗥𝘂𝗹𝗲𝘀 𝗼𝗳 𝗣𝗿𝗼𝗱𝘂𝗰𝘁 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗺
🤔 “𝗪𝗿𝗶𝘁𝗲 𝗙𝗶𝗿𝘀𝘁 𝗼𝗿 𝗕𝘂𝗶𝗹𝗱 𝗙𝗶𝗿𝘀𝘁? 𝗪𝗵𝘆 𝗔𝗜 𝗶𝘀 𝗥𝗲𝘄𝗿𝗶𝘁𝗶𝗻𝗴 𝘁𝗵𝗲 𝗥𝘂𝗹𝗲𝘀 𝗼𝗳 𝗣𝗿𝗼𝗱𝘂𝗰𝘁 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗺𝗲𝗻𝘁
This week I had the pleasure of listening to a presentation by Brent Smith, who highlighted the value of prototyping and empowering builders in the age of AI.
🔧 Why Prototyping Matters
Prototyping isn’t new—it’s a smart investment in any product development process. It enables early detection of design flaws, improves usability through real user feedback, and reduces costly mistakes before full-scale production. Prototyping aligns designs with manufacturing constraints, accelerates time to market, and builds stakeholder confidence by turning ideas into tangible, testable solutions.
Another super interesting conversation with Ilya Sutskever at the Dwarkesh Podca
Another super interesting conversation with Ilya Sutskever at the Dwarkesh Podcast[1]which made me fall short on my exercises in the gym last night as brain detached from body to process the input. Still on it.
One interesting aspect is Ilya talking about how essentially from his perspective it’s again time for research as “just” scaling the current set of technologies likely will generate a large amount of revenue with a considerable amount of cost, but research is required to get to the next level of AI.
This is a great start into the week. Creates some fast smiles and than some more serious afterthough
This is a great start into the week. Creates some fast smiles and than some more serious afterthoughts.
I absolutely agree that working backwards from real customer needs must be guiding product design. Not featuritis or Buzzword Compliance. No smart for the smart product label.
I also love technology which ease my life. I don’t mind a smart environment. The thing is - and I’m addressing my fellow technologists here - we need to make sure that we design those systems in reliable fashion.
I still remember how fascinated I was by ELIZA. It must have been in the mid-90s
I still remember how fascinated I was by ELIZA. It must have been in the mid-90s. From [1]:
“𝘌𝘓𝘐𝘡𝘈 𝘪𝘴 𝘢𝘯 𝘦𝘢𝘳𝘭𝘺 𝘯𝘢𝘵𝘶𝘳𝘢𝘭 𝘭𝘢𝘯𝘨𝘶𝘢𝘨𝘦 𝘱𝘳𝘰𝘤𝘦𝘴𝘴𝘪𝘯𝘨 𝘤𝘰𝘮𝘱𝘶𝘵𝘦𝘳 𝘱𝘳𝘰𝘨𝘳𝘢𝘮 𝘥𝘦𝘷𝘦𝘭𝘰𝘱𝘦𝘥 𝘧𝘳𝘰𝘮 1964 𝘵𝘰 1967 𝘢𝘵 𝘔𝘐𝘛 𝘣𝘺 𝘑𝘰𝘴𝘦𝘱𝘩 𝘞𝘦𝘪𝘻𝘦𝘯𝘣𝘢𝘶𝘮. 𝘊𝘳𝘦𝘢𝘵𝘦𝘥 𝘵𝘰 𝘦𝘹𝘱𝘭𝘰𝘳𝘦 𝘤𝘰𝘮𝘮𝘶𝘯𝘪𝘤𝘢𝘵𝘪𝘰𝘯 𝘣𝘦𝘵𝘸𝘦𝘦𝘯 𝘩𝘶𝘮𝘢𝘯𝘴 𝘢𝘯𝘥 𝘮𝘢𝘤𝘩𝘪𝘯𝘦𝘴, 𝘌𝘓𝘐𝘡𝘈 𝘴𝘪𝘮𝘶𝘭𝘢𝘵𝘦𝘥 𝘤𝘰𝘯𝘷𝘦𝘳𝘴𝘢𝘵𝘪𝘰𝘯 𝘣𝘺 𝘶𝘴𝘪𝘯𝘨 𝘢 𝘱𝘢𝘵𝘵𝘦𝘳𝘯 𝘮𝘢𝘵𝘤𝘩𝘪𝘯𝘨 𝘢𝘯𝘥 𝘴𝘶𝘣𝘴𝘵𝘪𝘵𝘶𝘵𝘪𝘰𝘯 𝘮𝘦𝘵𝘩𝘰𝘥𝘰𝘭𝘰𝘨𝘺 𝘵𝘩𝘢𝘵 𝘨𝘢𝘷𝘦 𝘶𝘴𝘦𝘳𝘴 𝘢𝘯 𝘪𝘭𝘭𝘶𝘴𝘪𝘰𝘯 𝘰𝘧 𝘶𝘯𝘥𝘦𝘳𝘴𝘵𝘢𝘯𝘥𝘪𝘯𝘨 𝘰𝘯 𝘵𝘩𝘦 𝘱𝘢𝘳𝘵 𝘰𝘧 𝘵𝘩𝘦 𝘱𝘳𝘰𝘨𝘳𝘢𝘮, 𝘣𝘶𝘵 𝘩𝘢𝘥 𝘯𝘰 𝘳𝘦𝘱𝘳𝘦𝘴𝘦𝘯𝘵𝘢𝘵𝘪𝘰𝘯 𝘵𝘩𝘢𝘵 𝘤𝘰𝘶𝘭𝘥 𝘣𝘦 𝘤𝘰𝘯𝘴𝘪𝘥𝘦𝘳𝘦𝘥 𝘳𝘦𝘢𝘭𝘭𝘺 𝘶𝘯𝘥𝘦𝘳𝘴𝘵𝘢𝘯𝘥𝘪𝘯𝘨 𝘸𝘩𝘢𝘵 𝘸𝘢𝘴 𝘣𝘦𝘪𝘯𝘨 𝘴𝘢𝘪𝘥 𝘣𝘺 𝘦𝘪𝘵𝘩𝘦𝘳 𝘱𝘢𝘳𝘵𝘺.”
☕ 𝗪𝗵𝘆 '𝗔𝗚𝗜 𝗶𝘀 𝗦𝘁𝗶𝗹𝗹 𝗮 𝗗𝗲𝗰𝗮𝗱𝗲 𝗔𝘄𝗮𝘆'? 𝗛𝗼𝘄 𝗦𝗵𝗼𝘂𝗹𝗱 𝗬𝗼𝘂 𝗦𝗵𝗮𝗽𝗲 𝗬𝗼𝘂𝗿 𝗔𝗜 𝗦𝘁𝗿𝗮𝘁𝗲𝗴𝘆
☕ 𝗪𝗵𝘆 “𝗔𝗚𝗜 𝗶𝘀 𝗦𝘁𝗶𝗹𝗹 𝗮 𝗗𝗲𝗰𝗮𝗱𝗲 𝗔𝘄𝗮𝘆”? 𝗛𝗼𝘄 𝗦𝗵𝗼𝘂𝗹𝗱 𝗬𝗼𝘂 𝗦𝗵𝗮𝗽𝗲 𝗬𝗼𝘂𝗿 𝗔𝗜 𝗦𝘁𝗿𝗮𝘁𝗲𝗴𝘆
Writing this early morning in Seattle with a much-needed cup of coffee - currently at an exciting ML conference and spent my long flight diving into this eye-opening conversation between Andrej Karpathy (ex-Tesla AI Director, OpenAI) and the amazing podcast host Dwarkesh Patel in the Dwarkesh Podcast [1]. His insights offer crucial guidance for tech and business folks navigating AI adoption.