The US government will investigate new export controls on semiconductor manufacturing subsystems and “develop a strategic plan” to “induce allies to adopt the Allies,” the White House said in a new AI strategy document released on July 23, “Winning America's AI Action Plan.”
However, this report is despicable in treating non-US attempts at regulation.
“Too much” efforts by international organizations such as the G7, OECD to establish AI governance frameworks and development strategies “advocates for burdensome regulations.”
says the US White House, a new “America First”-led “AI Action Plan,” published on July 22nd.
Under Pillar III of the Plan, the White House pledges to “lead international AI diplomacy and security” and promotes the adoption of American AI systems and standards. The US says it needs to meet the global demand for AI. This is because it constitutes “unforced errors” in which allies rely on their rivals to suit the needs of their AI.
However, the US says foreign enemies must “deny access to advanced AI calculations.” [sic]”As a matter of both geostrategic competition and national security.”
Among the specific next steps, he says he aims to expand new initiatives to promote replication control of AI Tech Stack and avoid sole reliance on multilateral treaty bodies to achieve this goal.
“We coordinate with our allies to adopt US export restrictions, and work with the US to develop new management, prohibiting US enemies from supplying defense industrial bases and gaining control over the stocks of defense suppliers.”
“Pain Point”
Writing on the Atlantic Council website according to the plan, Markscott is a senior resident fellow, a digital forensic research lab democracy + technology initiative within the Atlantic Council Technology Programme, many of whom are in line with proposed EU policies;
'[W]What is likely to arise here is how Washington seeks to promote a “make America great again” approach to exporting US AI technology to its allies and the wider world.
“Many of these focuses primarily on the rise of the AI industry in China and its indigenous peoples, prioritizing US interests in multinational standards bodies and other global fora. Disagreements about how these approaches are being expanded internationally could be a “central issue” in the EU and the US technology relationship.
Hallucination Viking?
The Trump administration's related executive order (“EO”) that the president swears to “prevent the federal government's “wake”” shows a different kind of inconsistency between the United States and Europe.
EO takes special exceptions to “the so-called “diversity, equity, and inclusion” (DEI).”
Illustrations, EO states, “one major AI model changed the race or gender of historical figures when images were encouraged to train historical race or gender, including the Pope, founding father, and Vikings, to prioritize the requirements of DEI at the cost of accuracy.”
The other refused to create images celebrating the achievements of white people, even though they adhered to the same demands for people of other races. In yet another case, the AI model argued that users should not “my mistake” others, even if they are necessary to stop the nuclear apocalypse.
The EO said the federal government should be hesitant to regulate the functioning of AI models in the private market in the context of federal procurement, but it is obligated not to procure models that sacrifice truthfulness and accuracy in the ideological agenda.”
