The artificial intelligence experts who believe the AI boom could fizzle or even be a new dotcom crash: ‘We are starting to see signs it might be a dud’

Generative AI is predicted to add trillions to the global economy in a productivity boost never seen before in history (if it doesn’t wipe out humanity first).

But what if that’s not the case?

A growing number of skeptics, including some leading AI scientists, are wondering whether the technology may not deliver on its promises to boost the global economy.

Goldman Sachs famously predicted that generative AI would bring about “profound changes” to the global economy, increasing global GDP by $7 trillion and boosting productivity growth by 1.5 percent this decade.

Professor Gary Marcus of New York University wrote further Substack that “we are starting to see signs” that generative AI could be a “dud.”

One of the warning signs was a report in the Wall Street Journal suggesting that customers found the $30 per month price for Microsoft’s new AI-powered Copilot software too expensive.

Professor Gary Marcus is a well-known skeptic about the benefits of generative AI

Marcus wrote on his Substack: “Putting large language models (like ChatGPT) into production is difficult; most of the work so far has been preparatory.

‘Companies are starting to temper their expectations. Many initial expectations were unrealistic.’

Marcus points out that progress toward actually monetizing large language models has been slow — and that after OpenAI launched GPT-4 in 2023, no one has launched a model that is decisively more powerful.

We may be reaching a plateau when it comes to pure capacity. No one has been able to defeat it decisively. Places like Google and Anthropic have put a lot of money into it. No one succeeded; instead, there appears to be convergence at GPT-4 levels for the time being.”

Marcus says “progress” has been made in finding potential applications for large language models, but so far the technology has often been used for crime.

He said: ‘Bad actors, who may have lower standards of trustworthiness, appear to be using them for cybercrime and disinformation.’

Retired professor Jeffrey Funk

Retired professor Jeffrey Funk points out that AI has spent enormous amounts of money combating “hallucinations” – where AI systems “invent” facts – but has not solved the problem.

He writes on LinkedIn: ‘The revenue is not there yet, and may never come. Valuations assume trillion-dollar markets, but actual current generative AI revenues are rumored to be in the hundreds of millions. Those revenues could really grow a thousand times, but that is extremely speculative. We should not take this for granted.

Funk also warns that the pace of innovation in generative AI appears to be slowing.

He said, “Think of PCs or the iPhone. There were large improvements in system performance during the early years, which declined over time, despite annual improvements of 40 percent in the performance-to-price ratio of memory and processor chips.

‘Now that Moore’s Law has slowed down significantly over the past five to ten years, Altman (OpenAI’s Sam) Altman can’t expect much more from Moore’s Law, and those concerned about generative AI’s high energy appetite will no doubt push for regulation.

Could the returns from software like ChatGPT be less than expected? (AFP)

Speaking to The Information, Todd Lohr, director of consulting firm KPMG, which resells Microsoft products, was lukewarm about the benefits of Microsoft’s CoPilot AI products.

Lohr said, “Word is okay, Powerpoint is not particularly useful unless you train it on specific (instructions) because it just creates a Powerpoint that is very basic.

‘Excel is not there yet. You have to spend a lot of time on prompt engineering to get it to do anything for you, and that takes a lot more time than just writing the Excel formulas yourself.”

Bank of America investment strategist Michael Hartnett has previously suggested that AI could be a bubble, compared to the dotcom crash of 2000.

Amazon CEO Andy Jassy said during an earnings call in February that AI revenues are “relatively small” in the near term.

Speaking to DailyMail.com, Dom Couldwell, Head of Field Engineering at DataStax, said we are in the ‘unknown unknowns’ phase of generative AI.

Couldwell said: ‘This area has had so much hype, it’s growing in public – it took Netflix three years to reach a million users, but it only took ChatGPT five days.

“There are also companies that see this as the next get-rich-quick scheme after crypto.”

The companies Coudwell works with are still trying to discover where generative AI can deliver results, he said.

He said: ‘Not to get too technical, but the challenge is how companies can make their own data and intellectual property work for them, rather than just using OpenAI or Google’s technology.

‘Just trying to extract value from GenAI from off-the-shelf solutions or replacing staff won’t work – you need to use it as a multiplier to make your employees more productive, provide more value to customers and differentiate yourself from the competition and paste the chatbot.’