In a stark warning to US lawmakers, former Google CEO Eric Schmidt argued that the ongoing artificial intelligence revolution is profoundly underestimated, declaring it “underhyped” Schmidt cautioned that neither policymakers nor the public are fully prepared for the seismic shifts AI is poised to bring to the nation and the world.
“The arrival of this new intelligence will profoundly change our country and the world in ways we cannot fully understand,” Schmidt testified, emphasizing the urgency and scale of the transformation. “And none of us… is prepared for the implications of this.”
Schmidt painted a picture of rapid advancement far beyond popular conceptions centered on chatbots like ChatGPT. He highlighted the imminent development of AI programmers capable of replacing traditional software engineers and AI mathematicians rivaling top graduate students – developments he expects within the next year. “What it really is, is a reasoning and planning system that we’ve never seen before,” he explained, stressing the fundamental nature of the technology.
This technological leap, however, comes with unprecedented demands, particularly for computational power and energy. Schmidt underscored the massive infrastructure requirements, pointing to industry plans for “10 gigawatt data centers.” To put this in perspective, he noted, “An average nuclear power plant in the United States is one gigawatt.” This illustrates the sheer scale, leading to projections that data centers could consume an additional 29 gigawatts by 2027 and a further 67 gigawatts by 2030.
“These things are industrial at a scale I have never seen in my life,” Schmidt stated. He called for urgent action to boost energy production across all forms – renewable and non-renewable – and highlighted significant regulatory hurdles, such as the average 18-year timeline for power transmission approvals, which need federal preemption to be expedited.
Beyond energy, Schmidt identified the need for high-skilled immigration to fuel innovation and “light touch regulation” focused specifically on cyber and biological threats posed by AI misuse.
Perhaps the most critical concern raised by Schmidt was the race towards “superintelligence” – intelligence surpassing human capabilities. With some industry estimates placing its arrival within a decade, following human-level AI possibly within three to four years, Schmidt stressed the paramount importance of the US reaching this milestone first.
He pointed to China’s rapid progress, citing the emergence of competitive models like DeepSeek despite US efforts like chip restrictions. “They’re clever and they’re smart. They have industrial programs, huge grants going into these companies,” he warned. “Welcome. China has arrived into the competition.”
Schmidt framed this competition in stark national security terms, arguing it overshadows even geopolitical flashpoints like Taiwan. “If they come to superintelligence… first,” he cautioned, “it changes the balance of power globally in ways that we have no way of understanding, predicting, or dealing with.”
Here is Schmidt’s speech in its entirety:
I’m here to tell you that I honestly believe that the AI revolution is under hyped, and here’s why. The arrival of this new intelligence will profoundly change our country and the world in ways we cannot fully understand. And none of us, including myself and frankly, and anyone in this room, is prepared for the implications of this.
What’s happening at the moment in our industry is that we’re very, very quickly, for example, developing. Developing AI programmers and these AI programmers will replace traditional software programmers we’re building in the next year, AI mathematicians that are as good as the top level graduate students in math.
This is happening very quickly. You can look at this in a number of the, uh, products today. You think of AI as chat GPT, but what it really is, is a reasoning and planning system that we’ve never seen before. The implication of this is profound. Uh. In terms of the way the algorithms work, they’re gonna need a lot more computation than we’ve ever had.
They’re gonna need a lot more energy, and I’ll talk about that. What does the industry need? We need high skills. Immigration, we talk to you about this every day. Uh, light touch regulation around cyber and bio threats. We can talk about that. And most importantly, we need the energy and the numbers are profound.
What we need from you, if, if I may say that directly is need energy in all forms — renewable, non-renewable, whatever, it needs to be there and it needs to be quickly. I and others are investing in things like fusion, which are incredible, but they’re not going to arrive soon enough for the need. And I’ll frame this at the end by my comments about China.
So people are planning 10 gigawatt data centers. Now, just to do the translation. An average nuclear power plant in the United States is one gigawatt. How many gig, how many nuclear power plants can we make in one year where we’re planning this 10 gigawatt data center gives you a sense of how big this crisis is.
Many people think that the demand in of energy part that our industry takes will go from 3% to 99% of total generation. One of the estimates that I think is most likely is that data centers will require an additional 29 gigawatts of power by 2027 and 67 more gigawatts by 2030. Gives you a sense of the scale that we’re talking.
These things are industrial at a scale I have never seen in my life. The, in the terms of energy planning, the current model is mostly natural gas peaker plants plus renewables. And that’s probably gonna be the path we’re gonna have to follow, right? To get there. And, uh, for all the reasons that you can imagine.
We have a bunch of regulatory issues around fixing the, the energy grid. It takes on average 18 years to get the power transmissions and so forth to put these things in place. We need to find federal ways to preempt that and make it happen faster in order to deal with the needs. Many of these data centers, by the way, are in the heartland.
They have a huge economic impact positively on areas that hip typically do not have the kind of growth that they would like. Now, why is this all important? When you build these systems, you have intelligence in the computer, and then eventually human level intelligence. Some people think it’s within three to four years.
Then after that you have something called super intelligence. And super intelligence is the intelligence that’s higher than of humans. We believe as an industry that this could occur within a decade. It is crucial that America get there first. What is China doing? They’re leading in something. Go open source.
I. They’re very close behind us. You all have done a great job in doing chip restrictions and things like that to try to slow them down. They’re clever and they’re smart. They have industrial programs, huge grants going into these companies, and they’re weaponizing up in the sense of competition. If you look at deep seek, deep seek showed up, right?
Nobody expected this. It turns out it’s on par now with some of the top models. Welcome. China has arrived into the competition. What would happen if China beat us? I. Let’s think about it. The path to intelligence, that’s superhuman intelligence. Think of the national security implications of that competition.
This is why I believe, and I’ll say directly to you, that although everyone is concerned about Taiwan, I’m much more concerned about this because if they come to super intelligence, the strong form of intelligence first. It changes the balance of power globally in ways that we have no way of understanding, predicting, or dealing with.
Thank you, Mr. Chairman.