Latest / DEEP INSIGHT / Americas Oldest Bank Bets Big On AI

Transcript

Americas Oldest Bank Bets Big On AI · DEEP INSIGHT

• 0:01
  Speaker 1: At America's oldest bank, one hundred and thirty four. New
• 0:04
  Speaker 1: workers don't sleep, they don't take sick days, they don't
• 0:06
  Speaker 1: even have names. They're what b and Y calls digital employees.
• 0:11
  Speaker 2: Our digital employees have their own log in. They communicate
• 0:14
  Speaker 2: with their human managers via teams or email.
• 0:18
  Speaker 1: B and Y says, they're not the same as AI agents.
• 0:21
  Speaker 2: So our digital employees are slightly different from agentic solutions,
• 0:25
  Speaker 2: which are like freelancers who come in and do a
• 0:27
  Speaker 2: defined scope of work, execute that brilliantly, but leave.
• 0:31
  Speaker 1: So how do you communicate with the digital employee when
• 0:34
  Speaker 1: they mess up? Rachel Lewis manages nine digital employees, and
• 0:39
  Speaker 1: she showed me what it looks like on a day
• 0:40
  Speaker 1: to day basis.
• 0:41
  Speaker 3: In our payments face we support the settlement of like
• 0:44
  Speaker 3: two and a half to three trillion dollars a day.
• 0:47
  Speaker 3: So we can see here the number of repairs of
• 0:49
  Speaker 3: digital employees and to take in so far the average
• 0:52
  Speaker 3: time it's taking them to process these transactions, and most importantly,
• 0:58
  Speaker 3: you know how many repairs for successful.
• 1:00
  Speaker 1: In this instance, the payment was missing a country code.
• 1:02
  Speaker 1: The digital employee identified the problem and fixed it a
• 1:06
  Speaker 1: year ago. That would have been done by a human,
• 1:08
  Speaker 1: but it wasn't a role people like doing.
• 1:10
  Speaker 4: It's really focus on very specific, repetitive tasks that really
• 1:15
  Speaker 4: allow our human employees to do much more human, intense,
• 1:20
  Speaker 4: interesting type roles.
• 1:21
  Speaker 1: B and Y employees forty eight thousand humans, down from
• 1:24
  Speaker 1: about fifty three thousand a few years ago. The company
• 1:27
  Speaker 1: is building more digital employees, but it's not as simple
• 1:30
  Speaker 1: as AI replacing humans. They're upscilling all of their employees
• 1:34
  Speaker 1: at the same time.
• 1:35
  Speaker 2: We really thought that the most important thing for us
• 1:39
  Speaker 2: was to have a workforce who was AI literate.
• 1:42
  Speaker 1: To do that, B and Y developed a platform that
• 1:44
  Speaker 1: pulls in a variety of AI models that are integrated
• 1:47
  Speaker 1: with the firm's internal data and compliance. They named it Eliza,
• 1:51
  Speaker 1: a tribute to the wife of their founder, Alexander Hamilton,
• 1:54
  Speaker 1: America's first Treasury Secretary. Ninety nine percent of BNY's employees
• 1:59
  Speaker 1: have completed a ten now Eliza training. I got the
• 2:02
  Speaker 1: ten minute version. Okay, we're going to create a new agent.
• 2:06
  Speaker 2: For personal use. There's your very first agent, Peggy.
• 2:10
  Speaker 1: We asked Peggy to help prepare some questions and do
• 2:13
  Speaker 1: some research.
• 2:14
  Speaker 2: Maybe you could say, give me a brief history of
• 2:17
  Speaker 2: b and Y.
• 2:18
  Speaker 1: I cut to the chase and ask Peggy who in
• 2:20
  Speaker 1: the financial services industry is doing the best job in AI.
• 2:24
  Speaker 2: It's fired if it doesn't say Bmy, it didn't say
• 2:27
  Speaker 2: bmway double performance to view for Eliza today.
• 2:31
  Speaker 1: Well, at least we know Peggy's not biased. While it's
• 2:34
  Speaker 1: probably too early to say whether or not B and
• 2:37
  Speaker 1: Y is the best at AI, we do know that
• 2:39
  Speaker 1: the firm spent nearly four billion dollars on technology in
• 2:42
  Speaker 1: twenty twenty five, about nineteen percent of its total revenue.
• 2:46
  Speaker 1: That's the highest proportion of its large bank peers.
• 2:50
  Speaker 5: There's an AI arms race. The banks are part of that.
• 2:53
  Speaker 5: But you don't define success by who spends the most.
• 2:56
  Speaker 5: You define success by who has the best results and why.
• 3:00
  Speaker 5: And almost every bag does not disclose benefits from AI,
• 3:04
  Speaker 5: and I sometimes wonder whether it's BE and Y or
• 3:06
  Speaker 5: the industry. They even know their ROI the returner on investment.
• 3:11
  Speaker 5: It's a lot of spraying and praying when it comes
• 3:14
  Speaker 5: to spending on tech.
• 3:15
  Speaker 1: Generally, Goldman Sachs's research team screen for companies in the
• 3:18
  Speaker 1: Russell one thousand that would be the biggest beneficiaries from
• 3:21
  Speaker 1: AI due to high labor costs that could be automated.
• 3:25
  Speaker 1: B and Y ranks toward the top of the list
• 3:27
  Speaker 1: with nineteen percent earnings upside, and that's why B and
• 3:31
  Speaker 1: Y is going all in. You guys are in this
• 3:34
  Speaker 1: intense AI session for four weeks, which is super cool.
• 3:38
  Speaker 1: This is B and Why's AI boot Camp, forty hours
• 3:40
  Speaker 1: of training mostly for non engineers, and the goal find
• 3:44
  Speaker 1: ways to automate their jobs who can't participant a shop.
• 3:47
  Speaker 1: Cherion is an attorney for the firm, working to streamline
• 3:50
  Speaker 1: his day to day.
• 3:51
  Speaker 6: The thing I'm working at the boot camp. The project
• 3:54
  Speaker 6: is to create an agent that can generate a contract
• 3:57
  Speaker 6: negotiation playbook, and that generally used to take a like
• 4:00
  Speaker 6: sixty seven hours just to generate, so now it's becoming
• 4:03
  Speaker 6: you can do within thirty minutes.
• 4:05
  Speaker 1: And Why is aiming to put fifteen thousand of its
• 4:07
  Speaker 1: employees through boot camp this year. It currently has one
• 4:10
  Speaker 1: hundred and sixty AI solutions in production.
• 4:13
  Speaker 2: We can certainly help build technology and engineering, but the
• 4:17
  Speaker 2: great ideas are going to come from the practitioners in
• 4:19
  Speaker 2: the business.
• 4:20
  Speaker 1: But I kept coming back to the obvious questions. Where
• 4:23
  Speaker 1: the boot campers huddled in a lab together unknowingly coding
• 4:26
  Speaker 1: away their future job security. How many humans can and
• 4:30
  Speaker 1: will be replaced by digital employees which work twenty four
• 4:33
  Speaker 1: to seven.
• 4:34
  Speaker 5: I wouldn't think about it that way.
• 4:36
  Speaker 4: I would think about it more as unlocking that productivity,
• 4:40
  Speaker 4: enabling all employees to be productive.
• 4:42
  Speaker 1: Everyone I spoke with it be and why envisioned a
• 4:44
  Speaker 1: future of the firm that still relies on humans alongside machines.
• 4:48
  Speaker 1: And this is the first time the centuries old company
• 4:51
  Speaker 1: has had to contend with new technology. Maybe I should
• 4:54
  Speaker 1: have asked Eliza what she thinks, Assass