AI’s Transformative Power: Key Takeaways from Harvard’s Leading with AI Conference
Last month, I participated in Harvard Business School’s Leading with AI: Exploring Business and Technology Frontiers conference. This event brought together the latest thinking around artificial intelligence from the Digital Data Design Institute at Harvard. My mind was blown with all the new ideas that were shared at this conference. I am now spending more time learning about artificial intelligence (AI), Large Language Models (LLMs), and different AI tools to better understand how these may impact our day-to-day lives.
My key takeaways from the conference are:
1. AI is transforming everything
Multiple speakers likened the AI transformation to the Internet revolution and the creation of the steam engine. Some quotes from panelists who are working closely with AI tools highlight how much they think this will change our ways of working:
“AI is the single largest value creation opportunity in the history of mankind. This is in terms of economic value, longevity, and quality of life.” — Marcelo Claure, Founder & CEO, Claure Group
“AI is going to be your colleague.” — Ethan Mollick, Associate Professor, Wharton School of the University of Pennsylvania
“All of Michael Porter’s Five Forces can be transformed with AI. We all have to immerse ourselves in what’s changing.” — Mary Erdoes, Chief Executive Officer, Asset & Wealth Management, JPMorgan Chase
Professor and Author Ethan Mollick raising AI questions for organizations
2. AI can support greater efficiency, personalization, and access
AI for Efficiency
One of the biggest areas of opportunity is efficiency. AI technology is evolving rapidly. For example, ChatGPT-4 is now scoring almost 100% on tests like GRE, SAT, LSAT, and the Bar Exam. Studies also show that it is raising work performance — workers who used AI to support their work were able to do 12.5% more work, 26% faster, and with 40% higher quality. More data on these studies is in Ethan Mollick’s book, Co-Intelligence: Living and Working with AI.
Some examples shared include:
Marketing: Sir Martin Sorrell predicted that media buying is going to be disrupted.
Healthcare: Sham Kakade talked about how doctors can ask AI to search medical publications and provide a summary on medical questions
Music: Derek Ali and Bas shared how musicians can use AI tools to create beats and music.
AI for Personalization
The potential for AI in healthcare is significant. AI is scoring 85% on medical exams! Armen Mkrtchyan from Flagship spoke about how AI can support personalized healthcare and prevention. For example, long before someone is diagnosed with pre-diabetes, AI can identify the biomarkers that should be tracked and provide personalized recommendations. Additionally, a patient’s health history could be summarized to help clinicians make decisions.
However, several issues need to be addressed for personalization in healthcare to be truly effective:
More clinical data is needed in LLM systems for individual recommendations.
More studies need to include underrepresented populations to avoid biased data.
Insurance companies lack incentives to invest in preventative technology, raising the question of who would fund these advancements.
AI for Access
Just as the internet created global access to knowledge, AI has the potential to democratize knowledge. Khan Academy now has Khanmingo, which provides on-demand AI-powered support for education. Multiple speakers discussed how AI can enable people in less-resourced countries to access knowledge in transformative ways. Derek Ali and Bas also spoke about how AI tools can help musicians who don’t have access to sound equipment and engineering to still make music.
3. AI raises questions that society will need to grapple with
Like any new transformative technology, there are many critical questions that need to be addressed. When one of the conference co-chairs asked the audience if they think AI will lead to the end of our civilization, most people said no. However, the side effects of AI can be enormous, similar to the impact of technologies like social media on mental health.
Some key areas that need exploration include:
Regulatory Standards: How will this field be regulated? What kind of transparency is needed? How do you ensure artists get paid for their work?
Managing Biases: AIs are based on the knowledge they are trained on, which is inherently biased depending on the data used for training.
Mitigating Misuse: There are many ways that AI can be used for malintent, particularly in the realm of impersonation. During the conference, they shared samples of AI video technology to show how easily someone can be impersonated which rases ethical and security issues.
4. Leaders need to invest in learning about AI now
To understand the potential and limitations of AI, leaders need to spend time using the tools available. Some anecdotes that have stuck with me:
Ethan Mollick says people need to spend a minimum of 10 hours using AI to understand its potential and limitations.
Karim Lakhani shared, “HBS will have a new required class for first-year MBAs on AI.”
Mary Erdoes said, “The winners will be management teams that immerse themselves in this — you can’t outsource this. Leaders need to make this part of their daily work.”
5. Some AI tools to explore
Professor Mitch Weiss sharing AI tools to get started
During the conference, I took note of various AI tools that were mentioned. Here are a few I plan to explore:
ChatGPT-4, CoPilot, Gemini, Claude: Currently the most popular for general consumers. People use these to create draft documents, synthesize text, upload photos to analyze, map data, create personas, and more. Note the paid version yields stronger results than the free version so its worth upgrading for regular use.
Beautiful — for presentations
Generative AI studio — for images
Jarvis — for code review
Midjourney — for images
Zapier — for customer support
We also had an amazing session about creativity in AI with music led by industry leaders. I will write another post about that, but here are some music tools that were shared:
Clearly, it’s an exciting and nerve-wracking time to be part of the early days of AI. I don’t think we should shy away from it, but instead, immerse ourselves to fully understand the benefits and potential harms.