Just for fun, I asked ChatGPT: What do INFORMS members mean when they use the phrase 'artificial intelligence'?
Original Message:
Sent: 05-22-2024 12:41
From: Rahul Saxena
Subject: Member OPPORTUNITY of the Week: AI and Your Work
Thanks, Irv, for posing the question of "what does INFORMS mean by AI".
AI is very broadly defined. When asked for a 1-sentence definition, ChatGPT says: AI is the simulation of human intelligence processes by machines, especially computer systems.
The decision intelligence (DI) subset of AI is more germane to INFORMS. Gartner says: DI advances decision making by explicitly understanding and engineering how decisions are made and how outcomes are evaluated, managed and improved via feedback. Ref https://www.gartner.com/en/information-technology/glossary/decision-intelligence.
The INFORMS vision aligns with DI: Better decision-making for a just, prosperous and sustainable world. Interestingly, the "AI Factory" of Iansiti and Lakhani works as the capability for DI. In this article from 2021 I explored this topic a bit further: https://rahul-saxena.medium.com/narrow-ai-for-decision-intelligence-ada9e0d01001.
Since LLMs have become salient, I think all of us would have tried them out and started to use them. In my "craft" work, LLMs provide content that I can use after tweaking, especially summaries, draft papers, and illustrations. The "mass-produced" decision support analyses from our SaaS are most interesting for the outliers and surprises we unearth, for which we find the LLMs lacking.
We're exploring how to use LLMs to provide a conversational interface as a parallel to the web-browser interface. Looking forward to reading Segev Wasserkrug's article.
------------------------------
Rahul Saxena
FrogData.com
Original Message:
Sent: 05-22-2024 09:03
From: Irv Lustig
Subject: Member OPPORTUNITY of the Week: AI and Your Work
Rahul's response to the question raises another interesting question, which is "What do we (in INFORMS) mean when we use the phrase 'artificial intelligence'?"
Tracy was looking for input "from multiple members on how artificial intelligence is impacting their work."
Rahul used a definition from Iansiti and Lakhani that defined AI as "a computer system able to perform tasks traditionally handled by people".
With the latter definition, many of us could say that we have been working on various methods to be embedded in such computer systems, and thus we, as INFORMS members, have been doing AI way before it has become a mainstream topic. And members of our community are using methods such as machine learning, reinforcement learning, etc., to solve problems, and those methods are typically considered in the domain of "artificial intelligence" as defined by the computer science community.
I am hazarding a guess that what Tracy is really looking for is how "generative artificial intelligence" is impacting our work. Some interesting work in this area has been led by Segev Wasserkrug of IBM that was the topic of a talk at the recent INFORMS Analytics Conference. See https://meetings.informs.org/wordpress/analytics2024/speakers/segev-wasserkrug/ for a description. I believe he and other members of the INFORMS Practice Section will be publishing an article in OR/MS Today about their work.
------------------------------
-Irv Lustig
Optimization Principal
Princeton Consultants
Original Message:
Sent: 05-21-2024 06:52
From: Rahul Saxena
Subject: Member OPPORTUNITY of the Week: AI and Your Work
Hi Tracy,
Until January 2020, I didn't say I was working in AI. Then an article, "Competing in the Age of AI" by Iansiti and Lakhani, was published in the Harvard Business Review. It defined AI as "a computer system able to perform tasks traditionally handled by people". With this definition, I found that I had been working in AI, and that I was running an "AI Factory".
At present, the FrogData.com Decision Analytics SaaS that I work on provides decision advice and operations analyses to US car dealers to make a range of business decisions spanning sales, inventory, service, accounting, and HR. For example: which cars to buy, what price to set for each car each day, etc.
This system converts decision intelligence from craft to mass production. It makes decision intelligence accessible and affordable to an entire market segment. Affordability comes because of repeatability and automation. The costs of making enterprise data warehouses and analytics algorithms are amortized over hundreds of customers. System operations and support costs are minimized by job automation that lowers costs.
My work consists of running and improving this AI factory, while adding and supporting a growing number of decision-makers.
------------------------------
Rahul Saxena
FrogData.com
Original Message:
Sent: 05-15-2024 13:53
From: Tracy Cahall
Subject: Member OPPORTUNITY of the Week: AI and Your Work
This week, instead of offering a tip, I want to present an opportunity to you. As you may know, we frequently feature member-written articles in our magazine, OR/MS Today. Currently, we're working on a piece intended to incorporate insights from multiple members on how artificial intelligence is impacting their work. We'd like to include a wide range of perspectives from members of different ages, career stages (including students), academic departments, and industries.
If you're interested in sharing your thoughts on this topic, we invite you to reply to this thread with a 100-200 word summary detailing how AI has changed or is currently changing your work. By posting here, you're granting permission for OR/MS Today editor Kara Tucker to publish your words, along with your name. Please post your responses by May 31.
Alternatively, if you prefer talking to writing, you can share a video of no more than 60 seconds on the same topic.
We eagerly await your perspectives!
------------------------------
Tracy Cahall
Member Engagement Manager
INFORMS
Catonsville MD
------------------------------