Reimagining Our Future with Intelligence

A team of researchers created a summer program at Dartmouth College in 1956 with this time-bound objective: “An attempt will be made to find out how to make machines use language, form abstractions and concepts, solve kinds of problems reserved for humans, and improve themselves” (McCarthy et al., 2006). That’s a long time ago for something that should’ve been a 3-month project.

There has been more investment and excitement in the last five years than in the entire history of AI. With all the recent hype, we are seeing investors and customers show disappointment in the value of AI. AI has made significant progress and breakthroughs over time, but it has always suffered and benefited from the hype of possibilities.

I am a big believer in AI’s potential because there is no other technology capability that has the promise to boost human intelligence. We are going to see AI achieve groundbreaking feats in our lifetime. However, in the business context and especially for emerging markets, we must focus on tangible AI value. We must avoid burning our money and energy (brain cells and data consumption) on massive AI models that yield minimal value. Stuart Russell, a top AI leader, writes in Human Compatible, “Machines are beneficial to the extent that their actions can be expected to achieve our objectives.”

This should guide our thinking on AI for Business. We must first define the objectives we want AI to achieve. We’ve spent time thinking about this at Davu and came up with a “Reimagine Society with AI” perspective that looks across industries and sectors and lays the foundation for no-brainer AI outcomes we can aspire to. This clarity can help ensure we build beneficial AI (safe, valuable, scalable).

3 things we need to do to build Beneficial AI

1. Focus on AI Outcomes: Our Reimagine Society with AI shows an industry view to consider and champion so we can collectively select AI outcomes that can power growth and transformation. For instance, in Financial Services, we can reimagine credit scoring to drive more equitable lending practices (Kanaparthi, 2024). In Healthcare, we could reimagine diagnostic tools to drive improved early detection of diseases (S. Oh et al., 2024). We are working with stakeholders to actively experiment on use cases to drive prioritization, custom model needs, and data sharing. We encourage interested parties to reach out to reimagine@davu.ai if they want to shape AI outcomes for businesses and, by extension, society.
2. Focus on Value: Businesses need to focus on tangible business value. Often, large models require high computing cost, energy consumption, and tons of data, and yield low returns (societal benefit and monetary). Unfortunately, this is where most of the hype is today. Skip the hype and focus on what you need to generate real value. GenAI is cool, but does that application need more or faster content? What problem is it solving and for whom? There are meaningful use cases beyond GenAI, and we need to invest in them.
3. Focus on Simplicity: AI platforms need to make models easy to design, build, deploy and consume. During my studies in AI design and build, I quickly realized how powerful AI could be, but also how complex it was to use. AI requires a lot of data engineering and science, but we will never achieve massive benefits if business owners and non-technical people can’t use it.

Clear outcomes, tangible business value and simplicity of AI development will drive greater adoption, business growth and innovation. These are the keys to reimagining our future with intelligence.

‍