White House Tech chief wishes AI wasn’t used for the worst

Arati Prabhakar, the Director of the White House Office of Science and Technology Policy, was a key player in the Biden administration’s work with AI. An artificial intelligence executive order in 2023 was signed to reign in companies developing the tech, ensuring that it’s safer as it grows.

The technology chief was one of the first to show the president ChatGPT but has her own views on the technology. While clearly impressed by its prowess, Prabhakar appears a little disturbed by how quickly it was used for nefarious purposes.

Speaking with MIT Technology Review, Prabhakar said that the rise of deepfakes and sexual abuse images that stem from AI is “horrific”. While work is still being done to curtail this use of AI, Prabhakar does say that those managing payment processors could be doing a lot more work.

As AI image generation tech improves with each passing month, Prabhakar wants companies like PayPal to take more action against misuse. Monthly payments to image generators are usually handled by large payment processors, and it appears that not enough work is being done to counteract bad actors in the space.

She also says that it’s time for face recognition technology to be used in less “horrific, inappropriate uses”. For instance, wrongful arrests have been on the rise. A major incident with the arrest of Randal Reid in 2023, saw the man wrongfully jailed despite never being in the state after being falsely identified with facial recognition.

However, Prabhakar does see an upside in the AI rat race. Of course, a huge proponent for the USA, she claims that the US has several advantages over its “strategic” rival, China.

China’s advances in AI have led the US to impose export bans on certain hardware, including Nvidia’s consumer-grade graphics card, the RTX 4090. This led Nvidia to develop the 4090D to skirt around the new rules.

The concern is trying to figure out how to keep “an open door for people who come seeking America’s advantages”, at the same time as protecting any “critical assets like our intellectual property.”

A huge part of this success now seems to ride on the back of the CHIPS Act. Introduced by the Biden administration, it allocates funding to secure America’s chip-making future. As part of this, Intel has been awarded $8 billion to build new facilities and fabricators across the country.

Prabhakar states that America is in a unique position now that it has gone through. Outside of Intel, Prabhakar says that “TSMC, Samsung, SK Hynix, and Micron” are all planning on building in the US. Having all five would be a definitive advantage for the US, as currently, the Taiwanese-based TSMC is the most used and popular fab plant in the world.

TSMC provides facilities for companies like Nvidia and AMD to build vital components for their hardware. Companies like SK Hynix are best known for their RAM production, including VRAM which will be found in graphics cards.

As hardware becomes more important to the AI industry’s struggle to stay on top, fabrication plants will become more of a hot commodity in the US.

Featured image: GetArchive, Pexels

The post White House Tech chief wishes AI wasn’t used for the worst appeared first on ReadWrite.