AI, Part 2: AI Explains Itself on Business and Legal Landscapes
As an experiment, I thought it would be interesting to build off of Part 1 discussing the technical capabilities of generative AI. I asked ChatGPT 4o, Open AI’s “newest and most advanced model,” to write a post discussing AI’s potential business use cases and legal landscape.
Below is the prompt and response, both verbatim, with a few minor exceptions. I inserted Clock Speed G.C.’s response to key sections where ChatGPT 4o’s response was incorrect or incomplete. The text in quotes is unchanged (i.e., not text written by Clock Speed G.C.), with the exception of removing text regarding the provision of legal services and some minor formatting changes, such as font and text size.
Overall, ChatGPT 4o created a good output, although it was incomplete and lacked currentness in some cases. The general consensus for this and other experiments seems to be that it can provide an impressive first draft. Read on to see if you would answer differently. (Note that this was for experimental purposes. This post is for informational purposes only and should not be relied upon for legal or other advice. Any comments or editorials are general, high-level observations and not intended to be exhaustive. Please be sure to read the Disclaimer below.)
###
The Prompt: “Write a blog post from the perspective of a technology law firm about generative AI that has an introduction paragraph; a section on top three business uses cases with example companies; and a section explaining the government action at the US federal level, in California and other states, and in Europe. Be sure the introduction paragraph references the prior post explaining the technical functionality of generative AI.”
The Result:
“Embracing Generative AI: Legal Perspectives and Practical Applications
In our previous post, we delved into the technical functionality of generative AI, elucidating how this groundbreaking technology operates through advanced algorithms and neural networks. Today, we shift our focus to the tangible business applications of generative AI, showcasing its transformative impact across various industries. Additionally, we will explore the evolving regulatory landscape at the federal level in the US, in California and other states, and in Europe. ”
[ClockSpeed G.C. Comment: Not bad. No notes.]
“Top Three Business Use Cases of Generative AI
Generative AI is revolutionizing numerous sectors, providing innovative solutions and driving efficiencies. Here are three prime business use cases:
1. Content Creation and Marketing
Generative AI is increasingly used in content creation, enabling companies to produce high-quality, engaging material at scale. For example, OpenAI's GPT-4 powers tools like Copy.ai and Jasper, which assist marketers in crafting compelling copy for blogs, advertisements, and social media. These tools save time and resources while maintaining a high standard of creativity and relevance.
2. Product Design and Development
In the realm of product design, generative AI facilitates the creation of novel and optimized designs. Companies like Autodesk use AI to generate a plethora of design options based on specified parameters, leading to innovative solutions that might not be conceived through traditional methods. This approach is particularly beneficial in industries such as automotive and aerospace, where design efficiency and innovation are paramount.
3. Healthcare and Drug Discovery
Generative AI is making significant strides in healthcare, particularly in drug discovery and personalized medicine. Firms like Insilico Medicine and BenevolentAI leverage AI to predict molecular structures and generate potential drug candidates. This accelerates the drug discovery process, potentially leading to faster and more cost-effective development of new treatments.”
[Clock Speed G.C. Comment: Some notes. The priority in terms of significance to humanity seems to be missing from the models from which ChatGPT 4o is pulling. For example, Google recently announced the launch of the AlphaFold3 server, a project of Google DeepMind and Isomorphic Labs, which can predict and model structures composed of proteins, DNA, and RNA (among others) (link here). The output may be better served including this example and put healthcare higher up on the list. Notably, healthcare did not appear as a result in earlier versions of this experiment.]
“Government Action and Regulatory Landscape
As generative AI continues to evolve, so too does the regulatory framework governing its use. Below, we examine key regulatory actions at various levels.
US Federal Level
At the federal level, the US government is actively addressing the implications of AI through several initiatives. The National AI Initiative Act of 2020 aims to coordinate AI research and development across federal agencies. The Federal Trade Commission (FTC) has also issued guidance on the ethical use of AI, emphasizing transparency, accountability, and fairness to prevent discriminatory practices and protect consumer rights.”
[Clock Speed G.C. Comment: More notes: Currentness and completeness. In October 2023, President Biden issued a comprehensive executive order directing a sweeping review and implementation of AI security policies and procedures across the US Federal government (link here). In April 2024, President Biden followed on the first Executive Order with steps to protect AI’s use in government agencies and throughout the US (link here). This includes proposed revisions for standards and security in the development and use of AI, workforce and civil rights protections for users of AI, research funding, and government workforce development (among others). Additionally, on May 15, 2024, Senate Majority Leader Chuck Schumer announced a bipartisan roadmap to guide legislation and policy to accelerate AI innovation while addressing risks such as discrimination and election interference. (CNN article here and roadmap available here).]
“California and Other States
California remains at the forefront of AI regulation in the US, with laws such as the California Consumer Privacy Act (CCPA) that set stringent data privacy standards. Additionally, the California Privacy Rights Act (CPRA) enhances these protections, giving consumers more control over their personal information. Other states, including Virginia and Colorado, have enacted similar privacy laws, reflecting a growing trend towards robust data protection measures across the country.”
[Clock Speed G.C. Comment: Missing critical updates, including Colorado’s recent adoption of the Colorado AI Act, which puts Colorado out in front of California and other states. The Colorado AI Act takes a risk-based approach to regulating AI similar to the European Union (The Hill article here). Notably, when the input prompt states “Colorado” instead of “California,” the output still did not reference the Colorado AI Act. Additionally, in 2024, California lawmakers have introduced a series of proposed laws and regulations not mentioned in the ChatGPT 4o output. California’s proposed protections include examples such as requiring notice to consumers, testing of AI models prior to introducing them into products in the market, prevention of bias, and opt-outs for being included in training data. (For additional discussion, the International Association of Privacy Professionals has additional reading here).]
“Europe
In Europe, the regulatory landscape for AI is primarily shaped by the General Data Protection Regulation (GDPR), which imposes strict requirements on data processing and privacy. The European Commission has also proposed the Artificial Intelligence Act, aiming to establish a comprehensive regulatory framework that categorizes AI applications by risk level. This act seeks to ensure that AI systems are safe, transparent, and respect fundamental rights.”
[Clock Speed G.C. Comment: Completeness and currentness. Very important developments missing. On May 21, 2024, the EU completed the final approval for adoption of the AI Act (article here). While GDPR has significant impact on AI, the AI Act itself and recent adoption should receive the most of the attention in ChatGPT 4o’s response. This legislation will quite literally have global implications. More information on the AI Act in Europe is available here).]
“Conclusion
Generative AI presents remarkable opportunities for businesses, offering innovative solutions across various domains. However, the rapid advancement of this technology necessitates a robust regulatory framework to ensure its ethical and responsible use. [Removed text regarding provision of legal services.]”
####
Overall, ChatGPT 4o provides a reasonably strong first draft of the post. It does, however, lack completeness and currentness in essential areas. Notably, many of the developments and missing information are more than a month old, suggesting that sufficient time has passed for the models to incorporate the data in training.
For more information and insights in current developments and technology, follow Clock Speed G.C. and SpeedTake.
Disclaimer: The above post does not qualify as legal advice. Please be sure to read the Disclaimer here for full terms and conditions of use of this site. All sources last visited 5 June 2024.