Trump seeks to develop AI free of ideological biases
His administration is aiming to promote human development, economic competitiveness and national security.
Trump waving on the presidential plane / Mandel Ngan.
Last week, the Trump administration announced investment plans regarding artificial antelligence (AI) infrastructure in the United States. The Republican administration's intention is to maintain leadership in research and industrial innovation in AI.
As soon as he took office, Donald Trump signed an executive order to revoke certain policies and directives for U.S. innovation in AI, clearing the way for the United States to act decisively in order to retain world leadership in artificial intelligence.
To this end, the Republican administration argues that U.S. policy is to maintain and enhance AI dominance with the goal of promoting human development, economic competitiveness, and national security.
Goodbye to ideological biases?
The Trump administration is convinced that to maintain leadership, AI systems must be developed that are free of ideological biases or designed social agendas.
For this reason, the Stargate initiative (of which SoftBank and OpenAI are participants), announced by President Donald Trump to invest $500 billion is one of the Republican's big projects.
Why is neutrality in AI content important?
When using artificial intelligence, users are exposed to certain ideological biases that derive from public discourse, so these tools are perceived as "manipulative" rather than "neutral," which corrodes confidence in a technology destined to be of much use.
An article published in City Journal indicates some considerations to apply to minimize ideological distortion in AI tools:
- Prioritize accuracy and neutrality: carefully select training data.
- Advance interpretability research: interpretability tools that can pinpoint which parts of the training data influence specific responses could help identify and correct biases.
- Adopt transparency rules: legislation requiring AI providers to disclose specific data on training methods and data sources.
- Establish independent oversight: provide for independent, nonpartisan organizations to routinely evaluate AI models for political bias.
The new U.S. government is sure that it must improve its country's standing in the AI race. It won't be an easy task. But the Trump administration has already begun.
LATEST
