With its famous immobility, Bloomberg Terminal has long inspired devotion, bordering on obsession. Among traders, the ability to plan a strategy using software’s confusing volumes of numbers and text to break down information is the mark of a seasoned professional.
But as large amounts of data are fed into the Terminal—not just earnings and commodity prices, but weather forecasts, shipping logs, factory locations, consumer spending patterns, private loans, and more—valuable information is lost. “It’s become unacceptable,” said Shawn Edwards, Bloomberg’s chief technology officer. “You miss things, or it takes too long.”
To try to solve the problem, Bloomberg is testing a chatbot-style interface for Terminal, ASKB (pronounced ask-bee), built on top of a basket of different language models. The broader idea is to help financial professionals shorten labor-intensive tasks, and to make it possible to evaluate abstract investment ideas against data using natural language commands.
As of publication, ASKB’s beta is open to about a third of the software’s 375,000 users; Bloomberg has yet to specify a full release date.
WIRED spoke with Edwards at Bloomberg headquarters in London in early April. We discussed the momentum to revitalize the Terminal, even if traditionalists may fight the change, and Bloomberg’s efforts to dispel stereotypes.
The following discussion is edited for length and clarity.
WIRED: Shawn, tell me about the reason for this Terminal redesign.
Shawn Edwards: For years, Bloomberg has been adding to this dataset. Often, finding the right piece of data in a sea of information is the deciding factor of whether or not you succeed. It’s becoming more and more unacceptable: You miss things, or it takes too long.
The main problem we solve with generative AI is to help users find valuable information and integrate the world view into a specific vision.
The idea is that unused alpha is hiding somewhere in the data, and ASKB will help reveal it?
Yes. The user gets to ask a high-level question—the thesis in their head—instead of asking specific data points. ‘How will the war in Iran and changing oil prices affect my portfolio?’ That’s a big, big question with many dimensions. Can we assemble that answer in minutes?
In a situation where everyone has access to data, what will separate the average traders from the best?
These tools are not magic. They do not measure [employee] with great suddenness. The difference will be your ideas.
In the hands of professionals, it allows them to do a better analysis, a deeper research – sorting out 10 good ideas when they may only have one time. If you are a mediocre analyst, it will be 10 mediocre ideas.
Bloomberg suggests ASKB as a type of agent AI. On its face, it looks more like a chatbot interface than something that automates tasks. What is an agent about ASKB?
There are salaries that come out every quarter. My job as an analyst is to prepare for what might come out of that earnings call. For each company I prepare, I look at how their price compares to their peers, sift through a lot of documentation, look at their fundamentals, and so on. During lead time, I don’t sleep.
With ASKB, I can create workflow templates. I can write a long query, and say, ‘Hey, here’s all the data I’m going to need. Give me a glimpse of the bull and the bear, what is the road, what is the direction.’ Now, I want to plan [the workflows] or wake them up when I see this or that situation in the world.
