Large Language Models, or LLMs, caught the
world's attention in 2022. Along with achieving nearly perfect scores on standardized
tests, language translation tasks, and other benchmarks, LLMs are starting to increase
productivity and impact the workforce, as well. But these models didn’t just come out of the
blue. They’re the tip of the iceberg when it comes to artificial intelligence or AI. There
are thousands of applications of AI today, and the field has a long and fascinating history.
Ba
ck in 1997, for the first time in history, a computer beat the reigning world champion in
a chess match. The news spread like wildfire, and the computer, Deep Blue, achieved celebrity
status. On that day the ancient art of chess, having over 10 to the 40 possible games was taken
over by AI but chess might seem easy compared to the mind-boggling 10 to the 360 possible games in
Go, which was also overtaken by AI when AlphaGo beat one of the world's top Go players in 2016.
Today, we have even
more AI milestones. While LLMs like ChatGPT steal the spotlight, the impact of
AI goes way beyond that. AI powers self-driving cars that have driven approximately two billion
miles on the road. In medicine, AI is being used to help interpret medical images, like X-rays
and CT scans, and to identify patients at high risk for things like sepsis. Soon, it will be used
during surgeries to help guide doctors and predict outcomes. Web applications like Stable Diffusion
can transform simple text i
nto high-resolution images, and researchers are using AI algorithms to
significantly improve the drug discovery process by predicting molecular behavior. But wait,
there’s more! AI enables retailers to offer personalized product recommendations to customers
based on individual preferences. Manufacturers are turning to AI-powered robots to improve quality
control during production and even optimize delivery routes for lightning-fast delivery!
But how does AI work? The key is that computers le
arn from data with as little human intervention
as possible. In Chess and Go, for example, the computer can learn from millions of previous
games or even play against itself to improve. In surgical models, computers can analyze
millions of surgical videos and patient information and use that to advise doctors
during medical procedures. In every case, AI learns from existing data to make predictions
on new data it hasn’t seen before. One of the most popular ways to do this is by using neural
networks, which are models based on the way the cells in our brains communicate with one another.
But what about the future of AI? If history is any indication, AI will continue to learn
from data and help us on more and more tasks. New policies, both scientific and
social, will be drafted to keep AI safe, equitable, and aligned with our best interests.
More and more businesses will adopt AI in some way, to increase productivity, as more and more
scientists develop advanced AI agents that m
ay be deployed as chat-bots, physical robots, as part
of software, or in ways we haven’t yet imagined!
Comments