White House Proposes ‘AI Bill of Rights’
The blueprint, published Tuesday, includes five principles meant to “protect the rights of the American public in the age of artificial intelligence.” The first, “safe and effective systems,” emphasizes Americans’ right to safety from systems deemed ineffective by the creator or a third party, as well as those designed in bad faith. The second principle shuns algorithmic discrimination (i.e. “unjustified different treatment” based on demographic status). Two other principles assert individuals’ right to data privacy, as well as the right to know when an automated system is being used in a way that might impact personal outcomes. The last suggests individuals should be able to opt out of such systems and have access to “timely human consideration” in the event that a system produces an error.
According to the White House, these principles are designed to protect Americans’ civil rights and access to critical services, including healthcare, financial services, and safety resources. They’re also meant to help facilitate equal opportunities. There is mounting evidence that AI can be algorithmically biased due to the (often unintentionally) skewed data they’re based on. These biases have been known to negatively impact women and people of color in particular.
Not everyone is a fan of the proposed AI Bill of Rights. “There are some recommendations in the [AI Bill of Rights] that, if enacted into rules by policymakers, could handcuff America’s ability to compete on the global stage,” said the US Chamber of Commerce’s Jordan Crenshaw in a statement. (One could argue such competition decreases in value when it violates people’s privacy and results in a systemic loss of opportunities, but that might not be the Chamber’s stance.) Crenshaw went on to say that the Chamber’s bipartisan AI Commission would be publishing its own policy recommendations in early 2023 to help the US in its mission to lead the AI industry.
If the Chamber is really operating from a global perspective, the US is running behind. The European Union has been closing in on AI-regulating legislation since late last year, when the European Commission proposed the Artificial Intelligence Act. Though still in the process of becoming law, the AI Act would regulate automated systems according to their risk level. Canada has similarly been working on what it calls the Artificial Intelligence and Data Act, while Brazil advanced its Artificial Intelligence Bill to the Federal Senate last fall.
As of now, US companies are left to “regulate” their own AI-based systems without external oversight. The White House’s AI Bill of Rights requests that companies and government entities approach their systems’ outcomes with transparency so that the public can make more informed decisions.
Continue reading
AMD Buys FPGA developer Xilinx in $35 Billion Deal
The deal, which we discussed earlier this month, will give AMD access to new markets that it hasn't previously played in, including FPGAs and artificial intelligence.
Intel Launches New Xe Max Mobile GPUs for Entry-Level Content Creators
Intel has launched a new consumer, mobile GPU — but it's got a very specific use-case, at least for now.
AMD Has No Near-Term Plan For Hybrid Big, Little Cores on Same Silicon
AMD has no plans to build a hybrid CPU with big and little cores any time in the near future, but the company continues to evaluate the concept.
Massive ‘Kilonova’ Explosion Shows First Observed Birth of a Magnetar
We've never seen a magnetar come into being, but a new high-energy event several billion light years away might be the first — a kilonova that signals the merging of two neutron stars.