On Open AI’s Dev Day, Open AI revealed a new feature that had been slowly rolling out to beta testers, GPTs. Yes, that’s what they called it. GPTs. It’s an amazing new feature that improves the tool immensely, but they didn’t even ask it what it should be called, because a better term is Custom GPTs. Likely the reason for the name is to basically say that only GPTs can come from their (soon to be released) web store. And yes, like all other software taking advantage of a large userbase, its allowing for users to sell their custom made GPTs on Open AI’s soon-to-be-announced store. You can read more about here. (worth noting there are currently some security risks with using code interpreter function.) Since having access, we’ve been, busy. So far we’ve made about a dozen very different GPTs (Ugh, can we call them what they are? Transformers? At least that feels cool), and one which we’ll be going over here is practical, it’s a custom GPT provided the entire text of the latest AI Executive Order. This is important, because it shows us how the tool can “understand” regulatory guidance and extrapolate what that might mean. So, let’s get into what the thing can do, for east we called it AI EO Bot. (It was a letter short from having me tell it to write a parody of BINGO)
Last week we went into a brief bit about how the AI EO could impact Life Sciences and Workforce based on the White House Press Statement on the AI EO. For this project we’ve uploaded the entirety of the White House’s October 30th 2023 Artificial Intelligence Executive Order to the AI EO Bot’s Knowledge. Most of what you’re about to read is from the bot itself, it’s important to edit what comes out of ChatGPT, it shouldn’t make final products, it’s more a middle-editor for content. One fault of the GPT here is that it assumes the EO is legal, when it’s predominantly guidance for legal, an important nuance it doesn’t quite understand without additional tools, which we’ll have a module on next year.
As businesses incorporate more artificial intelligence (AI) into their operations, understanding regulatory frameworks becomes increasingly important. AI EO Bot is a specialized tool designed to help professionals interpret the AI Executive Order. It translates the complexities of AI policies into more digestible information, aiding compliance and implementation.
Capabilities of AI EO Bot:
– Focused Knowledge : AI EO Bot contains detailed information on the AI Executive Order, enabling it to offer specific guidance.
– Plain Language Translation : The bot simplifies the language of AI regulations, making it more understandable for a lay audience.
– Impact Analysis : AI EO Bot provides context on how the AI Executive Order may affect different aspects of business operations.
Benefits of AI EO Bot:
1. Clarifying AI Regulation : The bot clarifies aspects of AI regulations, aiding users in their understanding of these policies.
2. Aiding Compliance : By interpreting the AI EO into more straightforward terms, AI EO Bot supports organizations in adhering to guidance requirements.
3. Efficient Information Access : The bot offers a quick reference to the AI EO, helping save time and effort typically required to navigate complex documents.
AI EO Bot is a practical tool for those looking to better understand and follow AI regulations. It’s a supportive resource for businesses seeking to align their operations with current policies, ensuring that they are well-informed and compliant. As the landscape of AI governance evolves, such tools can be critical in helping organizations adapt and maintain regulatory compliance.
The utility of AI EO Bot extends significantly into the realms of life sciences, healthcare, and patient interactions, sectors that are particularly sensitive to regulatory compliance and ethical considerations.
For professionals working in life sciences, navigating the regulatory environment is as critical as the scientific research and development itself. AI EO Bot provides these professionals with a resource to quickly understand how AI regulations might impact their work, from research protocols to product development, ensuring that their innovations align with current guidelines.
In healthcare, where AI applications range from patient data analysis to robotic surgeries, AI EO Bot can serve as a reference point for healthcare providers and administrators. It helps in interpreting how AI regulations apply to patient privacy, treatment methods, and healthcare delivery services. This is particularly beneficial when considering the stringent requirements for compliance in handling patient data and the ethical use of AI in treatment scenarios.
Moreover, professionals working directly with patients can leverage AI EO Bot to stay informed about how AI Executive Orders might influence patient care. They can ensure their use of AI in diagnostics, treatment planning, and patient monitoring adheres to the highest standards set as of this EO. This not only helps in maintaining compliance but also builds trust with patients, who are increasingly aware of and concerned about the ethical use of their data and AI in their care.
By bridging the gap between complex legal text and actionable knowledge, AI EO Bot empowers life sciences and healthcare professionals to focus on their primary objective: improving and saving lives through innovation and compassionate care, all while staying within the bounds of regulatory mandates. AI EO bot is all still ChatGPT but focused on the AI EO, so it can also produce graphs, compare AI EO to web based information, or write a limerick about the AI EO en español. (sure, why not?)
En la Casa Blanca un día,
una orden de IA se hacía.
“Con ética y ley”,
el Presidente dirá,
y en la tecnología confía.
Okay, that’s enough about the tool in terms of what it capable of: here’s the tool itself!
Now lets talk about some important information I derived using the tool, because its nice to know about a tool, its better to see some results. First, a cool image to break up this monotonous text, this image is how AI EO Bot sees itself
Produced with Dall-E-3 using descriptions generated by AI EO Bot
The AI Executive Order outlines several key initiatives aimed at supporting workforce development in the age of AI:
1. Labor Market Analysis (Sec. 6(a)(i)) : A report will be prepared by the Chairman of the Council of Economic Advisers on the labor market effects of AI, offering insights into how AI is reshaping employment.
2. Federal Response to Workforce Disruptions (Sec. 6(a)(ii)) : The Secretary of Labor is tasked with reporting on the federal government’s capacity to support workers displaced by AI, assessing existing programs and identifying potential new support mechanisms.
3. Principles for Workplace AI (Sec. 6(b)) : Principles and best practices will be developed by the Secretary of Labor to guide employers on the use of AI in a way that protects and enhances employee well-being. This includes managing job-displacement risks, upholding labor standards, and handling AI-related data collection ethically.
4. Guidance on AI and Compensation (Sec. 6(b)(iii)) : The Secretary of Labor will issue guidance to ensure that AI-augmented or monitored work complies with the Fair Labor Standards Act, ensuring workers are compensated fairly.
5. AI Education and Workforce Prioritization (Sec. 6(c)) : The NSF Director is instructed to prioritize AI-related education and workforce development, supporting a diverse and AI-ready workforce through suitable programs and awards
The AI Executive Order sets forth several initiatives aimed at enhancing healthcare and patient care through AI:
1. AI Tech Sprints (Sec. 5.2(f)) : The Secretary of Veterans Affairs will host AI Tech Sprint competitions to foster the development of AI systems that improve veterans’ healthcare. These competitions will provide technical assistance and mentorship to participants, encouraging innovation in AI applications for healthcare.
2. HHS AI Task Force (Sec. 8.2(b)(i)) : The Secretary of Health and Human Services (HHS) will establish a task force to develop a strategic plan for the responsible deployment of AI in healthcare, public health, and human services. The plan will address areas such as healthcare delivery, financing, and public health.
3. Quality and Performance Standards (Sec. 8.2(b)(i)) : The task force will also work on policies and frameworks for AI-enabled technologies in healthcare, focusing on long-term safety, real-world performance monitoring, and equitable principles to mitigate discrimination and bias.
4. Data Privacy and Security (Sec. 8.2(b)(i)) : Safety, privacy, and security standards will be incorporated into the software development lifecycle to protect personally identifiable information in the health sector.
5. Local Implementation Strategy (Sec. 8.2(b)(iii)) : The Secretary of HHS will consider actions to ensure compliance with Federal nondiscrimination laws by healthcare providers using AI.
6. AI Safety Program (Sec. 8.2(b)(iv)) : An AI safety program will be established to create a framework for identifying clinical errors from AI in healthcare settings.
7. Drug Development Strategy (Sec. 8.2(b)(v)) : The Secretary of HHS will develop a strategy for regulating AI in drug development processes, ensuring appropriate regulation throughout the drug development phases.
These initiatives are designed to ensure that AI technologies are deployed in a way that improves healthcare quality, patient safety, and equity, while also fostering innovation and protecting privacy and security.
The AI Executive Order details future plans that will impact life sciences and biotech through the promotion of innovation and responsible AI system development:
1. National AI Research Resource (NAIRR) Pilot Program (Sec. 5.2(a)(i)) : The Director of the National Science Foundation (NSF) will launch a pilot program to integrate distributed computational data, models, and training resources for AI-related research.
2. NSF Regional Innovation Engines (Sec. 5.2(a)(ii)) : The NSF will fund and launch at least one Regional Innovation Engine focusing on AI-related work to address societal or workforce needs within 150 days.
3. National AI Research Institutes (Sec. 5.2(a)(iii)) : The NSF will establish four new National AI Research Institutes in addition to the current ones, within 540 days. (fun fact this was started back in May, although I’d imagine these would indicate four one in addition to the seven mentioned in the link, although if they don’t make four new ones, its worth noting there’s no real definition here of new.)
4. Training for AI Researchers (Sec. 5.2(b)) : A pilot program will be established by the Secretary of Energy in coordination with the NSF to train 500 new researchers by 2025 to meet the demand for AI talent. (Here’s the start of it, but clearly 16 million is currently banking on expanding existing programs, and womp womp, the EO fails to understand the full scope and impact of generative AI here, because combined with Deep Learning, a lot of programming education/training needs to be revisited. In most tech-integrated industries, new grads often require years of training to become competent on the systems any given company uses, and these new tools fundamentally change the needs of the curriculum.)
5. AI and Intellectual Property (IP) (Sec. 5.2(c)) : The USPTO Director will publish guidance addressing AI’s role in the inventive process and inventorship, and issue additional guidance on patent eligibility for AI and emerging technologies.
6. Mitigating AI-related IP Risks (Sec. 5.2(d)) : The Secretary of Homeland Security will develop a program to mitigate AI-related IP risks, including measures for enforcement, information sharing, and private sector assistance. Right now the DHS focus on AI is predominantly on surveillance. Which, really should be expected considering their mission.
Here’s the neat quote from that link to DHS though: “DHS will also clarify and modernize immigration pathways for such experts (AI Experts), including those for O-1A and EB-1 noncitizens of extraordinary ability; EB-2 advanced-degree holders and noncitizens of exceptional ability; and startup founders using the International Entrepreneur Rule.”
7. Healthcare Technology Development (Sec. 5.2(e)) : The Secretary of HHS will prioritize grantmaking and other awards to support the development of AI tools that advance personalized immune-response profiles for patients, improve healthcare-data quality, and promote health equity and researcher diversity through programs like AIM-AHEAD. (If it doesn’t count Lived Experience as the same value as Work Experience, it’s not going to have the intended effect, but let’s see how it goes.)
8. AI in Veterans’ Healthcare (Sec. 5.2(f)) : The Secretary of Veterans Affairs will host AI Tech Sprint competitions to improve the quality of veterans’ healthcare and support innovation in small businesses, providing technical and mentorship resources. (The link indicates they’re targeting burnout, so one day in, and already we’re off to mismanagement of the intention by using broad interpretations to pay for systems already planned for to improve competitiveness for medical staff retention.)
9. Regulatory Strategy for AI in Drug Development (Sec. 8(v)) : Within 365 days, the Secretary of Health and Human Services (HHS) is tasked with developing a strategy for the regulation of AI and AI-enabled tools in drug development. This strategy will:
– Outline the objectives, goals, and high-level principles for appropriate regulation at each stage of drug development.
– Determine areas that may require future rulemaking, guidance, or additional statutory authority.
– Assess current budget resources and personnel and explore the potential for new public/private partnerships.
– Evaluate risks associated with implementing section 4 of the order, which could involve the development and deployment of trustworthy AI systems.
Great, so HHS is likely to employ the CDC, FDA, and CMS to tackle this goal, NIH might get tapped for some aspects, but as it goes if you’re unfamiliar, CDC is public health, CMS is healthcare and insurance standards (for Medicare and Medicaid and essentially sets the bar of insurance), and FDA here is the drug approval process. FDA already has a Software as a Medical Device segment, since it got its bones in 2016, so that’s probably where we’ll see a bulk of guidance come from. The awkward part here is that this rule-making should probably be co-administered by DOD, due to the implications of generative AI, its ability to create new biological weapons, and the importance of strict boundaries on its use. While there is a fair amount on the improvements to Cyber defense there isn’t a lot of info on the AI EO as it pertains to the DOD relative to bio weapons with AI drug discovery, the vagueness of HHS involvement doesn’t exclude the possibility however.
Key Federal Agencies to Monitor:
National Science Foundation (NSF)
For the latest in AI research and development, the NSF is the go-to source. They are instrumental in the implementation of the National AI Research Resource (NAIRR), and their Regional Innovation Engines and AI Research Institutes are set to be hubs of innovation and collaboration.
Department of Health and Human Services (HHS)
HHS is the cornerstone for healthcare and pharmaceutical professionals. Not only does it provide strategic plans for AI in healthcare technology, but it also offers grant opportunities geared towards public health improvements and health equity.
Department of Energy (DoE)
The DoE’s training programs are creating the next wave of AI talent. Organizations interested in harnessing skilled professionals for AI advancements in biotech and pharmaceuticals should follow their updates closely. (Why the DoE? Let’s start with the fact that California’s Geothermal Brines use chromatography for lithium extraction, biofuels are always a conversation topic, and Agg is starting to integrate solar-use in concert with sensors, fermentation, it’s so much more than electricity.)
United States Patent and Trademark Office (USPTO)
As AI reshapes innovation, intellectual property protection becomes more complex. The USPTO’s guidance on AI-related patents is crucial for securing life sciences and biotech advancements.
Department of Veterans Affairs
Their AI Tech Sprint competitions are a breeding ground for healthcare innovation, offering a unique platform for companies to develop and pilot AI solutions.
Department of Homeland Security (DHS)
DHS is critical for companies concerned with AI-related intellectual property risks. Their strategies for mitigating AI-related IP theft are essential for protecting proprietary processes and data.
Council of Economic Advisers
Understanding AI’s impact on the labor market is crucial for strategic workforce planning. The Council’s reports provide insights that can help shape workforce development strategies.
How about a timeline for all of the agencies discussed?
Generated with Dall-E-3 on AI EO Bot
As the AI Executive Order ushers in a new era of compliance and innovation for life sciences, healthcare, and pharmaceutical sectors, understanding the timelines for each federal agency’s initiatives is crucial. This guide provides an overview of key agencies and the schedules they are working against, helping professionals stay on top of the changes.
Agency Timelines:
National Science Foundation (NSF)
– NAIRR Pilot Program : Within 90 days of the EO, the NSF will launch the NAIRR pilot program, essential for AI-related research and development.
– Regional Innovation Engines : At least one to be funded and launched within 150 days, focusing on AI-related societal and workforce needs.
– National AI Research Institutes : Establishment of four new institutes within 540 days, adding to the existing ones.
Department of Health and Human Services (HHS)
– Regulatory Strategy for AI in Drug Development : A strategy to be developed within 365 days, defining regulation throughout drug development phases.
Department of Energy (DoE)
– Training for AI Researchers : Aiming to train 500 new researchers by 2025, to meet the AI talent demand.
United States Patent and Trademark Office (USPTO)
– AI Inventorship Guidance : Guidance to be published within 120 days addressing inventorship in the AI inventive process.
– AI and IP Guidance : Additional guidance on AI and IP intersections to be issued within 270 days.
Department of Veterans Affairs
– AI Tech Sprints : Two 3-month competitions to be hosted within 365 days to improve veterans’ healthcare.
Department of Homeland Security (DHS)
– AI-related IP Risks Program : Development of a program within 180 days to mitigate AI-related IP theft risks.
Council of Economic Advisers
– Labor Market Report : A report on AI’s labor market effects to be prepared and submitted, although a specific timeline post-EO is not provided.
By keeping track of these timelines, professionals in the health and pharma sectors can align their strategies with the upcoming regulatory changes and innovations. These agencies are not just shaping the future of AI in health and life sciences, but also setting the pace at which this future arrives. Stay informed and plan ahead to leverage the full potential of AI in your field.
What you just read are some of the outputs from the tool, however, I’ve edited it, added some commentary, and additional links for contextual reference.
Now that you’ve seen what it can do, join us for our Zoom Next Month! We’re going to go over creating this GPT, and playing with it to get it to work how we wanted to.
The tool can do more, try it out, but remember it’s GPT running the show here, so it can also produce Dall-e-3 images (provided Dall-E-3 is working at the time, it does have a tendency to get bogged down due to demand), browse the web for outside information, and incorporate all of that, so play around with the tool, we’ll have a blog post next week dedicated to how this GPT was made, but to be honest, if you’re looking to build a Custom GPT the built in GPT Builder will help with instructions for the custom GPT, however if you’re incorporating uploads, then there’s a lot more to consider, which is what we’re going to go over next week!
Important to remember! AI EO Bot is still just as finicky as any GPT, so your wording matters, if you’re interested in learning how to better prompt, be on the look out for our public courses or contact us today for training! You can reach us at Learn@promptingsolutions.org
We’re holding an upcoming training at and with Open Biopharma Research and Training Institute, 100% of the proceeds of the training will be donated to Open Biopharma. Open Biopharma is dedicated to reducing the cost to produce biopharmaceuticals with the goal of reducing the price to patients. Be on the look out for this unique training opportunity!
Prompting Integration and Consulting LLC 