Naveen Zutshi is the CIO of Databricks.
There are two schools of thought when it comes to AI. The first is fear that AI will become “sentient” as the technology gets “smarter,” eventually augmenting human behavior or even taking over human tasks and jobs. The other, and more accurate, belief is that AI will act nothing like us. Why? Because AI is inherently not human.
Yes, today’s most commonly used AI tool, large language models (LLMs), can learn and match patterns. Yet, they struggle to catch their own mistakes or capture chain-of-thought reasoning. Rather than focus on the fearmongering of whether or not AI will “take over,” we must turn our attention to how AI’s greatest strengths are dramatically improving innovation for creating real societal impact in ways humans would never be able to.
Bioengineering: Unlocking Disease Prevention And Uncovering New Cures
AI has been a tremendous boon for bioengineering, from protein structure prediction to drug discovery and genetic sequencing.
Every protein has an intricate 3D structure; deciphering them is wildly expensive and time-consuming. Finding a single protein’s structure can take years of lab work. With that in mind, Google DeepMind and Isomorphic Labs developed AlphaFold 3, an AI model that uses pattern recognition, multiple sequence alignments, physical modeling and confidence scoring to predict protein structures in hours instead of years.
Today, over 2 million researchers in almost 200 countries use the AlphaFold Protein Structure Database to assist their research for tasks like fighting malaria, discovering osteoporosis earlier and making progress in new Parkinson’s treatments.
Medical Imaging: Shining A Light On The Most Critical Cases
Beyond bioengineering, AI-enabled triage tools can analyze thousands of X-rays, MRIs and pathology slides to help radiologists prioritize the most critical cases that require immediate attention. This is an important step in alleviating overworked medical professionals in light of the severe radiologist shortage.
Pentavere, a healthcare AI and data science company developed by DARWEN AI and a Databricks customer, similarly uses data intelligence to amplify the impact of workers in the medical field. By automating data abstraction from unstructured electronic medical records, medical professionals can identify patients eligible for approved medications or other therapeutics faster than ever. What’s more, DARWEN AI’s predictive analysis can also be used to prevent cardiac arrest in heart failure patients and sight loss for visually impaired patients.
Climate Modeling: Improving Predictions To Save Lives
AI also has an important role to play in predicting climate patterns. For example, the Spherical DYffusion model developed by the University of California San Diego and the Allen Institute for AI has become a powerful tool for global climate predictions. With the power of machine learning trained on atmospheric and historical data across thousands of locations, it can help predict and allow people to prepare for hurricanes and droughts 25 times faster than current models.
Despite all the societal good that AI can unlock, some estimates predict that total energy spend could increase four times by 2030 due to data center energy demand. But AI can help with this, too: AI can improve energy efficiency by predicting solar panel output, ultimately reducing tech’s demands on our power grids. Models are also now being trained to analyze and detect high-carbon intensity on local grids in real time, automatically informing systems to switch to energy-efficient models instead.
Improving Our World With Greater Investment In Solutions
In the coming years, the core challenges of our generation will only prove more difficult to solve, and we’ll need better tools to face them. AI can analyze untold amounts of data, find patterns and surface solutions at speeds even the smartest humans can’t.
The most successful use cases for AI and opportunities where we’ll see a positive impact of the technology will be ones that have clear benchmarks for success and solid training data. In the wake of the rising trust in synthetic data, it’s important to remember that the real world and the people who call it home are ultimately going to be the key for AI to reach its full potential.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
