One of the major challenges in reporting on the new administration is that it is difficult to understand who is actually calling the shot. Take AI. In comments on world leaders at the AI Action Summit last week in Paris, Vice President JD Vance sets one view on where the Trump administration should go to in AI. And we are partners who choose other people – foreign and certainly foreign companies to expand their own use of AI. ”
He specifically speaks about the importance of America having its own advanced semiconductor design industry. Currently, the AI industry has desperate scrambling to ensure adequate procurement of advanced computer chips on which AI systems are run. They are essential technologies, and building them in the US is concerned about the advanced AI capabilities, the traditional war with China that dampens semiconductor imports, or the reuse of manufacturing. Regardless, it's a reasonable priority.
Vance's speech was bullish about AI. Based on what he said, I had expected that many of the features of AI policies during the Biden administration era would be in 2022 chips and science law, like concerns about the misinformation generated by AI. The development of the advanced American semiconductor industry, as well as the limited exports of advanced semiconductors, remained here. And so far, the chip project has been going very well. So I was hoping the administration would stick to it or even double down.
Sign up here to explore the big complex problems the world is facing and the most efficient ways to solve them. It was sent twice a week.
Trump officials may still do that. But because of the actions of the different arms of the new administration, it is now difficult to convey: Doge. Certainly, as we have heard, Elon Musk's so-called government efficiency systematically fires all probation employees in many government sectors. This includes most people in the government who were hired to tackle AI. This includes a majority of staff working to implement the Chips Act. With each Axios, “74 post-doctoral, 57% of tip staff are focused on incentives” and “67% of tip staff with focus on R&D” were scheduled to be fired. (The chances are that the administrator will reverse course or limit scope as it does not appear to occur yet.)
To be clear: this is not a targeted cut. The administration has not decided to reverse the course of chips. That's precisely in the process of firing all recent jobs across the widespread scope of government, Musk's team might accidentally fire all of the recent expertise accumulated around AI.
At the start of the Biden administration, no one outside Silicon Valley was talking about AI. Even in the midterm elections in 2022, we were still before Chat Grit. It has only been the last two years that AI has begun to turn the world upside down.
First there was ChatGpt, then scramble from all other major high-tech companies in the US, and with the blossoming of Dall-E 2 and Midjourney, as well as other open source AI art generators, they went out of business. Transformed visual content available on the internet. Then came Deepseek and other AI models that can “infer” but now there is deep research that many people say can replace research interns.
It all happened over two years, so all government employment focused on our new AI-powered world is not only working on chips, but also includes the National Institute of Standards and everyone. He's a new recruit. Technology's AI Safety Institute researches AI and tests the risk and capabilities of commercial AI models. And that means most of them are probation adoptions that are likely to be affected by Doge's rampage through the federal government.
This is a very stupid problem. If the administration wants to give up on the US advanced semiconductor industry, that's one thing. But, as Vance said last week, if the administration wants to have an advanced semiconductor industry to stay competitive with the most important technologies of the 21st century, it's ridiculous to give up on it as it hired most people. It's ridiculous that people have been working on it recently.
One way AI raises unprecedented challenges for governments is how quickly it moves. It's not the first technique to remove many jobs or dramatically change the way other jobs work. But the faster these transitions, the more lives they could be to be detrimental to Koichi, making it difficult for democratic mechanisms of surveillance and regulation to work. If AI is slowly happening and we are gradually trying to change the way we live over the next 30 years, we are pretty optimistic that we will have enough time to properly, learn and design the right laws.
However, AI is happening very quickly, and the US government is generally not that good at making things faster, but it is continually at the back. And Doge appears to be responsible for making it worse – even through malice, through carelessness. Fire people is much faster than hiring them. Freezing funds are much faster than knowing where to best use. However, using AI requires you to quickly accumulate expertise, act quickly, and stay at the pace of change in the private sector. Doge was originally conceived as an institution that facilitated it. But so far, they've made it difficult.
For me, it seems very likely that the world-communicating AI will arrive during current management. This could replace tens of millions of jobs overnight, representing a major shift in the concentration of our society's power, and could warn experts that they could go fatally. It's wrong. Vance is not wrong about the advantages. Mask believes there is a very realistic opportunity for AI to destroy humanity, but he is not wrong about his shortcomings. But they both need to be serious if they want their administration in a position to do this well.
