Blog Viewer

Welcome to the first in a series of what I hope to be quarterly updates tracking the exciting evolution of AI through the lens of Pharmacy General Intelligence (PGI). You've likely heard of Artificial General Intelligence (AGI), the concept of AI matching or surpassing human cognitive abilities.¹ PGI takes this a step further, focusing specifically on AI's potential to perform at or beyond the level of a skilled pharmacist. Imagine an AI agent seamlessly integrated into pharmacy workflows, handling tasks like medication verification, dose adjustments, and even patient counseling notes.

My goal is to provide a clear overview of the AI landscape, highlighting both the advancements propelling us towards PGI and the remaining hurdles. Expect insights into policy changes, industry trends, and technological breakthroughs. Because AI innovation moves at a rapid pace, my goal if for these updates will arrive quarterly. Your feedback is invaluable, so please share your thoughts if this is (or is not) helpful! 

Q1 2025: A Whirlwind of Change

The year kicked off with a flurry of activity in the AI space, marked by significant policy shifts, groundbreaking industry announcements, and rapid technological advancements.

• Policy Shifts:
• The newly inaugurated Trump administration swiftly repealed Biden-era AI policy safeguards, signaling a move towards deregulation.²
• A proposed congressional bill, the Healthy Technology Act of 2025, aims to authorize AI prescribing, potentially removing major legal barriers to AI clinical agents.³
• The Stargate Project, a 500-billion-dollar investment into building AI infrastructure for OpenAI in the United States, was announced.⁴
• Vice President JD Vance, speaking at the Artificial Intelligence Action Summit in Paris, advocated for minimal AI regulation, reflecting the administration's push for innovation.⁵
• These policy changes and proposed legislation together indicate a strong federal push to accelerate AI development in the United States. While federal changes are happening, State Board of Pharmacy and insurance reimbursement restrictions are still a large barrier.
• Industry Trends:
• The concept of AI agents, powered by advanced Large Language Models (LLMs), is dominating industry conversations. These agents aim to automate tasks by leveraging LLM capabilities to achieve specific goals.
• Y Combinator's recent "Requests for Startups" highlighted a strong focus on AI-driven solutions, particularly those involving AI agents.⁶
• However, Amazon's delay of its LLM-powered Alexa due to accuracy issues underscores the challenges of deploying AI in real-world applications, especially in high-stakes environments like healthcare.⁷
• Satya Nadella, CEO of Microsoft, discussed the impact of AI agents on the workforce, envisioning a shift from output generation to output management. This concept is very interesting when applied to PGI.⁸
• Technological Advancements:
• DeepSeek R1, an LLM developed by the Chinese company DeepSeek, made waves with its reported ability to achieve comparable performance to leading models at a significantly lower training cost.^9,10
• Elon Musk's xAI released Grok 3, claiming substantial improvements in benchmarks like math, science, and coding.¹¹

The Road to PGI: Key Challenges and Opportunities

While progress is undeniable, significant challenges remain. In my view, the most critical hurdle is providing LLMs with the right data. Patient data is dynamic, protected, and requires careful curation. Building an effective PGI necessitates:

• Establishing a reliable, comprehensive, and timely data flow, encompassing patient records, guidelines, and health system protocols.
• Filtering relevant information to avoid overwhelming the LLM and compromising accuracy.

The shift towards pharmacists "managing" AI-generated outputs, as suggested by Satya Nadella, appears to be a logical first step. This approach could enhance efficiency while adhering to regulatory and payment requirements by keeping a pharmacist in the decision-making loop.

Looking Ahead

Q1 2025 has set a dynamic pace for AI development, particularly in the context of PGI.  I'm eager to see what the next quarter holds. Thank you for reading and welcome your comments!

References:

1. Google Cloud. "What is Artificial General Intelligence (AGI)?" Google Cloud Discover, https://cloud.google.com/discover/what-is-artificial-general-intelligence. Accessed [Date Accessed].
2. Executive Order on Removing Barriers to American Leadership in Artificial Intelligence. (2025, January). The White House. https://www.whitehouse.gov/presidential-actions/2025/01/removing-barriers-to-american-leadership-in-artificial-intelligence/
3. Healthy Technology Act of 2025, H.R. 238, 119th Congress (2025): https://www.congress.gov/bill/119th-congress/house-bill/238
4. OpenAI. (2023). Announcing the Stargate Project. OpenAI. Retrieved February 20, 2025, from https://openai.com/index/announcing-the-stargate-project/
5. Transcript of JD Vance remarks at Paris AI Summit (2025,Feburary): https://gist.github.com/lmmx/b373b9819318d014adfdc32182ab17ff
6. Y Combinator. "Requests for Startups." ycombinator.com, https://www.ycombinator.com/rfs.
7. "Revamped Alexa AI Assistant Hit by Month-Long Delay." PCMag, https://www.pcmag.com/news/revamped-alexa-ai-assistant-hit-by-month-long-delay
8. Dwarkesh Patel. "Satya Nadella." Latent Space, DwarkeshPatel.com, February 19th, 2025. https://www.dwarkeshpatel.com/p/satya-nadella
9. Stanford HAI. (2025, February 13). How Disruptive Is DeepSeek? Stanford HAI Faculty Discuss China’s New Model. Stanford HAI. Retrieved February 20, 2025, from https://hai.stanford.edu/news/how-disruptive-deepseek-stanford-hai-faculty-discuss-chinas-new-model
10. Elie Bakouch, Leandro von Werra, and Lewis Tunstall. "Open-R1: a fully open reproduction of DeepSeek-R1." Hugging Face Blog, 28 Jan 2025, https://huggingface.co/blog/open-r1.
11. Grok 3 Release: The age of Reasoning Agents (2025, February 19th) https://x.ai/blog/grok-3