The Bloomberg terminal is getting an AI makeover, like it or not

Given the difficulty of solving it, the Bloomberg terminal has long inspired devotion, bordering on obsession. Among traders, the ability to chart a path through a program’s dizzying scrolls of numbers and text to isolate outlying information is the mark of a seasoned professional.

But as a greater amount of data is fed into the terminal—not just earnings and asset prices, but also weather forecasts, shipping records, factory locations, consumer spending patterns, private loans, and so on—valuable information is lost. “It’s becoming increasingly unacceptable,” says Sean Edwards, Bloomberg’s chief technology officer. “You miss things, or take too long.”

To try to address the problem, Bloomberg is testing a chatbot-style interface for Terminal, ASKB (pronounced Ask-bee), which is built on top of a basket of different language models. The general idea is to help finance professionals condense labor-intensive tasks, and make it possible to test abstract investment theses against data through natural language prompts.

As of the date of publication, the beta version of ASKB is open to about a third of the software’s 375,000 users; Bloomberg has not set a date for the full release.

WIRED spoke with Edwards at Bloomberg’s lavish headquarters in London in early April. We discussed the motivation behind the station’s renovation, whether traditionalists might balk at change, and Bloomberg’s attempts to get rid of the hallucinations.

The following conversation has been edited for length and clarity.

WIRED: Sean, tell me the rationale for this complete overhaul of the station.

Sean Edwards: For years, Bloomberg has continued to add to our comprehensive data set. Often times, finding the right piece of data in a sea of ​​information is the deciding factor in whether or not you succeed. It’s becoming more and more untenable: you miss things, or it takes too long.

The core problem we solve with generative AI is helping users find key insights and assemble a worldview around a given idea.

The concept is that untapped alpha lies somewhere in the data, and ASKB will help bring it to light?

Yes. The user should ask the high-level question — the thesis in their head — rather than asking about specific data points. “How will the war in Iran and the change in oil prices affect my investment portfolio?” This is a very big question and has many dimensions. Can we put this answer together in minutes?

In a scenario where everyone is able to wade through the interwoven data set, what will separate the average traders from the best traders?

These tools are not magic. They don’t make average [employee] Suddenly great. The difference will be your thoughts.

And in the hands of experts, it allows them to do better analysis and deeper research – to sift through 10 great ideas when they might not have had time for just one. If you are a mid-level analyst, there will be 10 mid-level ideas.

Bloomberg is promoting ASKB as a form of agentic AI. On the face of it, it looks more like a chatbot interface than something that necessarily automates tasks. What is a proxy in ASKB?

There are profits coming out every quarter. My job as an analyst is to be prepared for what might come on that earnings call. For each company I prepare, I research how their pricing compares to their peers, look through a lot of documents, look at their fundamentals, etc. During earnings season, I don’t sleep.

Using 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 summary of the bull and bear issues, what the street is saying, and what the guidance is. Now, I want to schedule [the workflows] Or I turn it on when I see this or that situation in the world.