The tech world has seen this movie before. Back in 2017, it was all about blockchain, and it seemed like it was going to change everything. From military intelligence to ice tea makers, it was everywhere. Now, in 2023, it’s all about artificial intelligence, and there’s been a huge surge in AI apps and hacks that some people think will revolutionize the world, or maybe even destroy it.
Even though the hype around AI is said to have fueled a rebound of the S&P 500, OpenAI COO Brad Lightcap is here to rain on the parade.
In a recent interview with CNBC, Lightcap shared his view that the impact of AI might be overhyped when it comes to making a real impact on businesses’ bottom line.
“I think the overhyped aspect is that it can instantly deliver significant business change,” he said, pointing out that CEOs can’t just ask AI to magically boost revenue growth or cut costs without any effort.
Lightcap’s more cautious perspective is becoming more common among AI industry leaders.
Yann LeCun, the Chief AI Scientist at Meta, also had a tempered view of AI’s potential at Meta’s Fundamental AI Research team’s 10-year anniversary. He compared today’s AI to “cat-level” or “dog-level” intelligence, noting that even after being trained on the equivalent of 20,000 years of reading material, AI systems still struggle with basic concepts.
Regulators are also treading carefully when it comes to AI. Christy Goldsmith Romero, Commissioner at the Commodity Futures Trading Commission (CFTC), emphasized the need for balanced AI regulation in financial markets.
While some experts believe AI could add trillions of dollars to the economy, others fear it could lead to the destruction of humanity.
Lightcap’s take is even more moderate than that of OpenAI CEO Sam Altman, who has been a big player in shaping AI regulation.
The journey from AI to artificial general intelligence (AGI) is more about incremental progress rather than drastic business transformations, according to Lightcap. For now, AI should be seen as a tool to enhance our capabilities, not replace them.
“I don’t know if it’s the most surprising thing per se, but one of the coolest things we see is ChatGPT acting almost like a sidekick in that regard,” he said. “Almost like a research assistant.”
*Edited by Ryan Ozawa.*