Biden's AI EO - Beginning of an Era?

The executive order (EO) on AI issued by the U.S. government has sparked a lot of debate and discussion. Whether you lean Left or Right, you may have different views on the EO’s implications and objectives. However, as a technology executive (or anyone living in this day and age), you should be aware of how the EO affects your industry, your company, and your life.

At a higher level, the EO aims to foster secure, safe, fair, equitable, and trustworthy AI for the benefit of Americans and the global community. It covers a wide range of topics, from defense and healthcare to education and civil rights. It also outlines the government’s vision and strategy for promoting innovation and adoption of AI within clear ethical and legal boundaries.

Below is my takeaway of key points from the EO that I feel impact my professional and personal life:

Defense: The U.S. government is investing heavily in AI for defense purposes, its probably fair to say that the US government saw this coming - as evidenced by the recent large contracts awarded to tech giants like Amazon, Microsoft, Oracle, and Google, and the testing of unmanned fighter jets. Clearly, AI will play a crucial role in enhancing the national security and military capabilities of the U.S.

Healthcare/Life Sciences: EO emphasizes the need for regulatory oversight and guidance for AI applications in healthcare and life sciences, such as medical research, drug discovery, and clinical trials. This is to ensure that AI is used for the public good and does not pose any harm or risk to human health and well-being, and no one with access to a computer (which is everyone) creates a biological weapon, intentionally or unintentionally.

Education and Public Schools: The potential of AI to transform education and public schools, and aims to leverage AI to improve learning outcomes, access, and equity. This also involves developing AI literacy and skills among students and teachers, and fostering a culture of curiosity and creativity. US wants to build a strong STEAM capabilities at early age and this could be the opportunity to do it.

Finance and Other Industries: EO acknowledges the existing regulations and standards for AI deployment in finance and other industries, and urges for continued vigilance and monitoring to ensure that AI is used safely and responsibly. This also includes addressing the challenges and opportunities posed by AI for financial stability, consumer protection, and market competition. While the industry application of AI is broad and potential is unknown yet but directionally we are all headed there.

Talent Development: With this EO, the government supports the development and retention of AI talent, both domestic and foreign, by streamlining immigration processes, facilitating access to visas, and providing upskilling and reskilling programs across government. This is to ensure that the U.S. has a diverse and skilled workforce that can drive AI innovation and adoption, and continue to sell the American dream to everyone wanting to pursue a career in the field.

Protecting American Jobs: The concerns and impacts of AI on the labor market, and measures to protect and empower the working class. This is to ensure that AI does not displace workers, but rather enhances their productivity and quality of life. There is obviously a narrative that is floating publically and in the government that AI is going to take away most of the blue-collar and several white-collar jobs. EO clearly gives power to the people and unionize to fight such initiatives by their employers.

Transparency and Self-Declaration: We all saw it coming, EO mandates that companies disclose and share relevant information and data about their AI systems, such as safety tests, performance metrics, and potential risks. This is to ensure that AI is used in a transparent and accountable manner, and that the public can trust and verify the AI systems they interact with. While this only applies to big programs and those that aim to transform mankind.

New Standards from NIST: The EO assigns the National Institute of Standards and Technology (NIST) the task of developing and establishing standards and best practices for AI, such as technical specifications, testing methods, and evaluation criteria. This is to ensure that AI is aligned with public safety and welfare objectives, and that the U.S. maintains its leadership and competitiveness in AI. These standards are not very clear, like most of the EO, details will follow soon but i see it as a positive sign like standardization of web APIs.

Content Authentication: The efforts to authenticate and verify the origin and integrity of AI-generated content, such as images, videos, and text. This is to combat the spread of misinformation and deception, and to protect the public from fraud and manipulation. While the private sector has tried watermarking as a way to create this transparency but it is inadequate and will require far more innovation in the field.

Privacy: The EO underscores the importance of privacy and data protection, especially for foundational models and training data that underpin AI systems. This involves collaboration between the National Science Foundation (NSF) and other networks and organizations to ensure that privacy is respected and preserved in AI development and use. Recent incident of NYT and OpenAi is just the beginning of organizations and eventually, people wanting to stay of the grid from AI capabilities.

Civil Rights: The issue of algorithmic bias and discrimination, and emphasizes the need for ethical and fair AI development and training. This is to ensure that AI respects and promotes the civil rights and liberties of all people, and that AI does not exacerbate existing inequalities or create new ones. This will be tough to follow but if models are trained with such purpose in mind, at least we can expect more fairness in society, maybe not absolute.

Innovation and Openness: This to me is the most encouraging development from the EO. Creating a culture of innovation and openness in AI, and provides resources and support for small and medium enterprises, startups, and inventors in the AI space. This is to ensure that AI is not monopolized by a few players, but rather fosters a diverse and vibrant ecosystem of AI actors and stakeholders.

Establishing America as an Authority: Creating multilateral agreements and partnerships with other countries and goverments to set and enforce standards and norms for safe and trustworthy AI. This is to recognize the global nature of AI, and to ensure that the U.S. plays a leading and influential role in shaping the future of AI. Every country is worried about the same things, the question is really to the people and citizens - expecting out governments to gover themselves but what better options do we have.

Winners and losers of the RACE:?

Technology giants have undeniably shaped the world for the past three decades, evident in the dominance of the top companies on the S&P. Today, every sector, from banking to healthcare to automotive, aspires to emulate the prowess of tech companies, recognizing the profound influence they wield over our lives.

The EO serves to solidify the position of tech companies as global leaders for the foreseeable future. As we navigate this era, I identify four categories of winners in this transformative sector:

1. Semiconductors: Accelerating Transformation for AI Impact

Contrary to a blanket assumption that all semiconductor companies will thrive in the AI era, the reality is more nuanced. The AI wave presents opportunities for semiconductor companies to pivot and create superior products. While NVIDIA initially rode this wave adeptly, recent developments across Intel, TSMC, Qualcomm, and tech giants such as Amazon, Google, and Apple will determine the winners. The pace of innovation within these companies must shift towards meeting the demands of Large Language Models (LLMs) and diverse use cases.

2. Software: Navigating the AI Landscape with OpenAI's Influence

OpenAI, not the first mainstream software, certainly boasts the most impactful contribution in a short timeframe. Established software giants like Google, Meta, and Amazon, having delved into AI for years, must now extend their focus beyond enhancing proprietary products. With OpenAI and Microsoft leading the way, these companies need to develop software and models that can be adopted by others, potentially disrupting their own offerings. Google, with its robust in-house AI capabilities, is strategically positioned to lead in this evolution.

3. Data: The Challenge for Data Companies in an AI-Driven World

The landscape for data brokers and enterprise data companies is poised for challenges. The vast data corpus used to train models like OpenAI's ChatGPT has become monumental, making it easier for enterprises to leverage raw data's power rather than outsourcing. Data companies must differentiate themselves through proprietary data, like Bloomberg or Experian, or by building niche LLMs/Data models that provide significant value. Additionally, the Executive Order will heighten scrutiny on data usage, impacting companies like Google and Meta.

4. Industry Adopters: AI's Promise of Transforming Every Sector

The true promise of AI lies in transforming every industry, redefining how we perceive and operate within them. Unforeseen shifts, such as banks without physical branches or retailers operating exclusively through apps, exemplify the transformative power of AI. Industry adopters and innovators in AI practices will reshape education, the military, customer service, and more, ushering in a new era of possibilities.

As we embrace the AI revolution, the tech industry stands at the forefront of innovation, ready to capitalize on the opportunities that lie ahead.

The EO on AI is a landmark document that sets the tone and direction for the U.S. government and society in the era of AI. It also has significant implications for technology executives, who need to understand and adapt to the changing landscape and expectations of AI. AI is a challenge and an opportunity for technology executives to demonstrate their leadership and vision in the AI domain and their respective industries.