The tech industry is in frantic GenAI PR control

https://www.nber.org/papers/w34836

You read that right. A massive new study just dropped, and 6,000 verified executives have confirmed that Generative AI has provided near-zero measurable benefits for their companies. Instead, they just got a massive price tag and a pile of structural headaches. It is not surprising to anyone actually doing the work. What is truly wild is the frantic PR damage control we have seen from Wall Street and tech vendors over the last 48 hours.

The reality check:
Let's look at the actual data. The National Bureau of Economic Research just published a landmark working paper called Firm Data on AI. They did not just run a cheap internet poll. They verified and interviewed 6,000 CEOs and CFOs across the US, the UK, Germany, and Australia.

The results are a complete reality check. Over 80% of these firms reported that AI has had absolutely zero impact on productivity or employment over the past three years. When you look specifically at labour productivity, 89% of managers said there has been zero change. Even the top executives who claim to be leading the charge are only using the tech for an average of 1.5 hours a week!

Moving the goalposts:
For the past two years, we have had to listen to tech billionaires aggressively promising that AI was going to completely automate white-collar work. Guys like Microsoft AI CEO Mustafa Suleyman, OpenAI's Sam Altman, and Anthropic's Dario Amodei have been running a relentless media tour predicting that jobs will be wiped out in 12 to 18 months. Elon Musk literally looked the UK Prime Minister in the eye and said no job would be needed.

As this 6,000-firm study proves, the grand corporate job replacement they promised simply is not happening. So what do the tech vendors do when their trillion-dollar illusion hits a peer-reviewed brick wall? They move the goalposts again. Now we are suddenly hearing excuses from analysts about the J-Curve and how it actually takes a decade for new tech to show results.

Blaming the users:
Right on cue, we are now seeing an absolute flood of news articles and sponsored reports spinning this failure into a skills gap crisis. The new, desperate narrative is that the AI software is flawless, but the workforce is just too uneducated to use it properly.

It is pure copium. Rather than admitting their multi-billion dollar tools are just glorified autocorrects that hallucinate facts and cannot handle complex legacy enterprise data, they blame the employees. Conveniently, this fabricated panic has created a massive grift where consulting firms and tech platforms get to sell useless prompt engineering courses and AI readiness certifications to workers desperately trying to appease automated HR filters.