# EBay Spurns GameStop’s Bid as ‘Neither Credible Nor Attractive’

An episode of [Bloomberg Intelligence](https://invisiblenetwork.ai/podcast/326301337), 12 May 2026, 22 min.

Audio: https://podtrac.com/pts/redirect.mp3/tracking.swap.fm/track/UVBrz8bN8aM2Xe47PEPu/traffic.omny.fm/d/clips/e73c998e-6e60-432f-8610-ae210140c5b1/7f534b8c-9e91-429a-b102-ae3c00090c6d/ef3b6165-18da-4e37-a76e-b4480112a5df/audio.mp3?utm_source=Podcast&in_playlist=0ad812bf-fd15-4012-a18c-ae3c00090c76

Watch Paul and Scarlet LIVE every day on YouTube: http://bit.ly/3vTiACF. Bloomberg Intelligence hosted by Paul Sweeney and Scarlet Fu -Cecilia D'Anastasio, Bloomberg Video Game Reporter, discusses EBay rejecting a $56 billion takeover offer from GameStop Corp. Chief Executive Officer Ryan Cohen, describing the unsolicited bid as “neither credible nor attractive.” EBay’s board turned down the offer due to “uncertainty” around the financing plan, operational risks, and GameStop’s governance, among other factors. -Lily Meier, Bloomberg Retail Reporter, discusses Under Armour. Shares sank after its guidance for this year missed Wall Street projections, weighed down by exiting a partnership with NBA star Stephen Curry and a financial hit from the conflict in the Middle East. Annual revenue will decline slightly, but it would have been little changed if the company still had the deal with Curry, the sporting-goods maker said Tuesday. Analysts on average projected a gain of about 2%. -Sid Philip, Bloomberg Chief Correspondent for Global Aviation, discusses the Bloomberg Big Take story: “Boeing Bets Comeback on Trump, China and an Elusive New Plane.” Boeing’s new CEO, Kelly Ortberg, must soon decide how to replace the workhorse 737 aircraft, as the company tries to regain public trust after a series of sometimes deadly accidents. He's also trying to land a big sales deal in China, with Trump's help. Both the sale and the new plane will be key to reviving the aerospace giant. -- Gautam Mukunda, Lecturer at Yale School of Management and Bloomberg Opinion contributor, discusses his Bloomberg Opinion column: “AI Isn’t Built for Bar Weather’s Black Swan Era.” AI models are matching or outperforming physics-based models in medium-range weather forecasting and are faster and require less computational infrastructure. These models struggle with extrapolating outside their training data, which can lead to failures when they encounter unusual weather patterns. See omnystudio.com/listener for privacy information.

## Transcript

[0:02] Speaker 1: Bloomberg Audio Studios, Podcasts, radio news. You're listening to the
[0:10] Speaker 1: Bloomberg Intelligence Podcast. Catch us live weekdays at ten am
[0:14] Speaker 1: Eastern on Apple, Cocklay and Android Auto with the Bloomberg
[0:17] Speaker 1: Business App. Listen on demand wherever you get your podcasts,
[0:21] Speaker 1: or watch us live on YouTube.
[0:24] Speaker 2: One of the stories, Paul, that's Scott, you and me
[0:26] Speaker 2: scratching our heads is eBay. Gamestop's bid for eBay, a
[0:31] Speaker 2: company that is four times its size. Ryan Cohen, of course,
[0:34] Speaker 2: making this very dramatic takeover offer for the company. eBay
[0:38] Speaker 2: has now rejected that fifty six billion dollar takeover offer,
[0:41] Speaker 2: calling it neither credible nor attractive. I think that pretty
[0:44] Speaker 2: much tells you what it thinks.
[0:45] Speaker 3: I did investment banking for a long time, and I
[0:47] Speaker 3: can get it very creative with capital structures, but I
[0:49] Speaker 3: have a hard time seeing how they're how they would
[0:52] Speaker 3: finance this thing.
[0:52] Speaker 2: All right, let's bring in Cecilia Denastasio. She is our
[0:56] Speaker 2: video game reporter here at Bloomberg.
[0:57] Speaker 4: Nice.
[0:57] Speaker 5: Great to see you, Cecilia.
[0:59] Speaker 6: Thank you for having me.
[1:00] Speaker 2: What was the thinking behind GameStop and making a bid
[1:04] Speaker 2: for eBay?
[1:05] Speaker 5: You know, we actually didn't get a.
[1:06] Speaker 6: Lot of insight into what CEO Ryan Cohen thought that
[1:09] Speaker 6: the proposed synergies could look like. He never really went
[1:12] Speaker 6: out and said, hey, we run a video game, you know,
[1:15] Speaker 6: retail company, but also we're doing really well and collectibles.
[1:18] Speaker 6: We're doing really well selling Pokemon cards. eBay also doing
[1:22] Speaker 6: really well selling Pokemon cards and baseball cards. We could
[1:25] Speaker 6: have heard that in any of the interviews that Ryan
[1:27] Speaker 6: Cohen gave, but he didn't actually really go out and
[1:31] Speaker 6: say that. In fact, when he was given questions about
[1:33] Speaker 6: what the combined company could look like, you often sort
[1:36] Speaker 6: of dodged.
[1:37] Speaker 3: So, I mean, what's the story with GameStop these days?
[1:39] Speaker 3: I mean, I'm not a gamer, thank goodness, most of
[1:43] Speaker 3: my kids are not gamers. But it's almost a meme stock.
[1:47] Speaker 3: But what's the real call in this company these days?
[1:49] Speaker 6: You know, it's interesting because GameStop could be doing a
[1:53] Speaker 6: lot worse than it is doing. You know, they shuttered,
[1:57] Speaker 6: you know, thousands of stores over the last couple of
[1:59] Speaker 6: years to be a big ubiquitous presence at your neighborhood
[2:02] Speaker 6: strip mall. And you know, gamers are increasingly purchasing their
[2:06] Speaker 6: wares in online marketplaces and not at these at these
[2:09] Speaker 6: physical storefronts, but GameStop is still hanging on. CEO Ryan
[2:13] Speaker 6: Cohen said, you know, we could be doing worse. It's
[2:15] Speaker 6: not a great business, but it's hanging in there.
[2:18] Speaker 5: We could be doing worse.
[2:19] Speaker 2: It's not a great line that you want to hear
[2:21] Speaker 2: from the CEO of a company stock that you happen
[2:23] Speaker 2: to own. So having said all of this, there's been
[2:26] Speaker 2: some reporting that perhaps Ryan Cohen just wanted to boost
[2:30] Speaker 2: the stock price of game Stop because there's some clause
[2:33] Speaker 2: I guess in one of his compensation packages that if
[2:35] Speaker 2: the share price reaches a certain level, he gets another
[2:38] Speaker 2: big payout.
[2:39] Speaker 5: Can you talk us through what's.
[2:40] Speaker 6: Going on there, It's really hard to say. If the
[2:43] Speaker 6: companies did end up combining, then it's possible that he
[2:46] Speaker 6: would have been able to receive that payout and he
[2:48] Speaker 6: would have been very happy. But also at the same time,
[2:52] Speaker 6: a lot of analysts pointed out that the way that
[2:53] Speaker 6: Ryan Cohen approached eBay wasn't exactly so friendly. You know,
[3:00] Speaker 6: so he it was an unsolicited bid, he didn't publicly
[3:03] Speaker 6: state what he thought the synergies could look like. When
[3:06] Speaker 6: eBay came back saying that the deal felt neither credible
[3:09] Speaker 6: nor attractive. I think that kind of speaks to this
[3:13] Speaker 6: idea that you know, maybe this wasn't so thoroughly thought out.
[3:19] Speaker 6: I think a lot of analysts are very very confused
[3:21] Speaker 6: about the tent.
[3:22] Speaker 4: You did a lot to improve that outlook there. Yeah,
[3:25] Speaker 4: I mean that what a disaster that was.
[3:26] Speaker 3: So do you think game Stop would pursue a proxy
[3:30] Speaker 3: fight and take it to that next step?
[3:32] Speaker 5: It's possible.
[3:33] Speaker 6: You know, Ryan is a very unpredictable executive, which is
[3:36] Speaker 6: part of what makes him really fun to cover. He
[3:39] Speaker 6: has He posted a manifesto on x formally known as
[3:42] Speaker 6: Twitter describing how he thinks American business people are often
[3:46] Speaker 6: taking a really risk averse approach to running their companies
[3:49] Speaker 6: and he's really trying to carve out a model for
[3:51] Speaker 6: something really different. So it's possible he would pursue the
[3:53] Speaker 6: proxy fight. It's possible also that this was a publicity stunt.
[3:57] Speaker 3: Do we know whose advisors are his investment banking advisors?
[4:01] Speaker 3: The only thing I saw was he's got a financing
[4:03] Speaker 3: letter from TD Securities, which again tells you that highly confident,
[4:07] Speaker 3: highly confident for twenty billion dollars forty years.
[4:10] Speaker 4: But we don't really know who other advisors are.
[4:13] Speaker 6: He's a highly autonomous individual.
[4:15] Speaker 5: Stay with us. More from Bloomberg Intelligence coming up after this.
[4:22] Speaker 1: You're listening to the Bloomberg Intelligence podcast. Catch us live
[4:26] Speaker 1: weekdays at ten am Easterned on apple Cockplay and Android
[4:29] Speaker 1: Auto with the Bloomberg Business App, Listen on demand wherever
[4:32] Speaker 1: you get your podcasts, or watch us live on YouTube.
[4:37] Speaker 3: Under Armour reported some numbers today. I didn't see them,
[4:41] Speaker 3: but they can't be good. The stock's down sixteen percent here.
[4:45] Speaker 3: But I know somebody who doesn't know what's going on here.
[4:46] Speaker 3: Lillly Myers, she's a retail reporter for Bloomberg News, joining
[4:49] Speaker 3: us live here in our Bloomberg Interactive Broker studio. What
[4:52] Speaker 3: did under Armour have to say that maybe the market
[4:54] Speaker 3: didn't go for.
[4:55] Speaker 7: Yeah, so I think under Armour disappointed investors. I think
[4:59] Speaker 7: that they're the revenue for next year or this current
[5:03] Speaker 7: year is not not as high as they expected, and
[5:06] Speaker 7: their earnings per share was also lower.
[5:08] Speaker 5: They also mentioned the Middle East conflict.
[5:11] Speaker 7: As impacting their sales and the split from the Steph
[5:14] Speaker 7: Curry brand hitting their bottom line.
[5:17] Speaker 4: My question with under Armour, I'm a Nike person, by
[5:19] Speaker 4: the way.
[5:19] Speaker 3: Full disclosure always happened, and I'm not switching, but it's competitive.
[5:25] Speaker 3: I mean, there's Nike out there body dust for those
[5:28] Speaker 3: of us in the know, we pronounce it that way, and.
[5:30] Speaker 4: I'm just not sure. And it's there's so many niche players.
[5:33] Speaker 3: I mean, the kids here at Bloomberg they were in
[5:35] Speaker 3: all different kinds of sinkers that I've never seen before.
[5:38] Speaker 4: How does under Armour fit in there, do you think?
[5:40] Speaker 8: Yeah?
[5:40] Speaker 7: I mean I think it's a really competitive space, like
[5:43] Speaker 7: you described, both on.
[5:44] Speaker 5: The footwear side and the apparel side.
[5:46] Speaker 7: You know, under Armour is part of this restructuring plan
[5:48] Speaker 7: that they've been going through recently, has been trying to
[5:51] Speaker 7: really slim down the number of skews that they sell,
[5:54] Speaker 7: you know, kind of streamline the products and also cut
[5:57] Speaker 7: back on their discounts.
[6:00] Speaker 3: So is this you know, I understand that it's really
[6:04] Speaker 3: a product driven business. You can't just crank out the
[6:06] Speaker 3: same sneaker year after year after year. I mean every
[6:10] Speaker 3: year it's like fat it is fashion. Do they have
[6:13] Speaker 3: the resources, they have the capabilities to really compete against
[6:16] Speaker 3: the I guess the bigger players.
[6:18] Speaker 7: Yeah, I mean it'll be interesting to see. The hard
[6:20] Speaker 7: thing with some of the retail innovation cycle is that
[6:23] Speaker 7: it takes like eighteen months, you know to start seeing
[6:26] Speaker 7: some of that new product actually hit the shelves. So
[6:28] Speaker 7: I'll be interested to see what they come out with.
[6:32] Speaker 3: You mentioned they.
[6:34] Speaker 4: Lost Steph Curry. That's a big name.
[6:38] Speaker 3: How do they think about using athletes as part of
[6:41] Speaker 3: their marketing promotion?
[6:42] Speaker 4: Strategy is a core to them? And how big of
[6:46] Speaker 4: a loss was like Steph Curry?
[6:47] Speaker 7: Yeah, so for Steph Curry, I mean that was a big,
[6:51] Speaker 7: a big deal. They had worked with the Curry brand
[6:53] Speaker 7: for a long time. That was a really essential part
[6:56] Speaker 7: of their basketball business. So this quarter they said that
[6:59] Speaker 7: for this coming year, the Curry brand separation made it
[7:04] Speaker 7: so that their revenue was down slightly. If they had
[7:06] Speaker 7: stayed with Curry, it would have been flat for the year.
[7:09] Speaker 5: So it'll be interesting to see, you know, if they.
[7:12] Speaker 7: Try and side more big names or what comes next.
[7:14] Speaker 3: I mean, there's kids coming out of college every year.
[7:18] Speaker 3: Are they going to be aggressive?
[7:19] Speaker 4: What did they say? Their strategy is there because it's.
[7:23] Speaker 3: A tried and true strategy, particularly for athletic where Nike's
[7:26] Speaker 3: seen it. I mean, Michael Jordan, you know, that's the
[7:29] Speaker 3: great example. But are they committed to that part that strategy,
[7:33] Speaker 3: do you think.
[7:34] Speaker 5: Yeah, that's a good question. I mean, they definitely work
[7:36] Speaker 5: with athletes.
[7:36] Speaker 7: I'm curious to see how that will evolve moving forward,
[7:40] Speaker 7: especially without Curry.
[7:43] Speaker 3: What's the uh, what's the tariff situation with some of
[7:46] Speaker 3: these shoot companies these days?
[7:47] Speaker 4: I mean is it? Do they even talk about it anymore?
[7:50] Speaker 3: Because I know a lot of the retail companies they
[7:52] Speaker 3: used to others as well, used to just give you
[7:54] Speaker 3: a number. It costs US five hundred million this quarter
[7:56] Speaker 3: or forty million this this quarter. Is still an issue
[7:59] Speaker 3: for these apparel companies.
[8:00] Speaker 5: Yeah, it's interesting. So it's switched a little.
[8:02] Speaker 7: Now they're talking about what their expected tariff refund will be,
[8:05] Speaker 7: so they're.
[8:05] Speaker 5: Baking that in or on the positive side.
[8:08] Speaker 7: I think, you know, there's a lot we still don't
[8:09] Speaker 7: know about what companies will get out of tariff refund,
[8:13] Speaker 7: if they'll get the full amount, what that process will
[8:16] Speaker 7: actually look like, how long it will take. But that's
[8:18] Speaker 7: the discussion now, is you know the upside from that refund?
[8:21] Speaker 3: Now companies have actually you have to go like file
[8:24] Speaker 3: and say I want a refund right now?
[8:26] Speaker 4: Has the Nike done.
[8:27] Speaker 3: It something like I'm thinking of the big big name
[8:30] Speaker 3: in your retail.
[8:31] Speaker 4: Has Nike actually done that for example?
[8:33] Speaker 7: Yeah, so a lot of major retailers have have sued
[8:36] Speaker 7: for tariff refunds, and I think it'll be interesting to
[8:39] Speaker 7: see as as they become more of a reality, how
[8:43] Speaker 7: companies will go about it.
[8:45] Speaker 3: There still really isn't a stre There isn't I guess,
[8:49] Speaker 3: a policy framework to get the money right.
[8:51] Speaker 4: They're still working on that.
[8:52] Speaker 7: From my understanding, I think it's it's it's still in process.
[8:56] Speaker 4: How much is if Nike gets money they do with it?
[8:59] Speaker 4: Do they give it back to me? Who paid a
[9:01] Speaker 4: tariff on my kicks?
[9:03] Speaker 9: Here?
[9:03] Speaker 4: We don't know that.
[9:04] Speaker 7: Yeah, I mean, I think that will be really an
[9:07] Speaker 7: interesting thing to watch. I think consumer reaction will be interesting,
[9:10] Speaker 7: you know, consumers who have had to face those higher prices,
[9:14] Speaker 7: but also the company that's haed to shoulder a lot
[9:16] Speaker 7: of it.
[9:18] Speaker 10: Stay with us.
[9:18] Speaker 4: More from Bloomberg Intelligence coming up after this.
[9:25] Speaker 1: You're listening to the Bloomberg Intelligence podcast. Catch us live
[9:28] Speaker 1: weekdays at ten am Eastern on Apple, Cocklay and Android
[9:31] Speaker 1: Auto with the Bloomberg Business app. Listen on demand wherever
[9:35] Speaker 1: you get your podcasts, or watch us live on YouTube.
[9:39] Speaker 3: All right, So President Trump leaving today for Chinese has
[9:41] Speaker 3: got a plane full of executives.
[9:43] Speaker 4: Including executive for Boeing, Yep and Boyden.
[9:46] Speaker 3: You think about a company that's leveraged to Chin it
[9:50] Speaker 3: a I always think of Apple, But Boeing's right there.
[9:53] Speaker 3: And we want to talk to Sid Philip about that,
[9:55] Speaker 3: Bloomberg Chief correspondent for Global Aviation, because this could be
[10:00] Speaker 3: one of those trips where somebody, you know, China announces
[10:03] Speaker 3: a gazillion plane deal and you.
[10:05] Speaker 2: Know, actually something that the CEO has kind of hinted
[10:09] Speaker 2: at already saying that you know, Trump's visit would be
[10:13] Speaker 2: quote a meaningful opportunity for us.
[10:15] Speaker 3: Yeah, Sid Philip, he does all this stuff for Blomberg News.
[10:18] Speaker 3: What are you expecting here on this trip for Boweing?
[10:20] Speaker 3: What do you think Boeing wants to get out of this?
[10:22] Speaker 8: So Boeing's looking to secure its first big order from
[10:25] Speaker 8: China in almost a decade, and so we understand it's
[10:29] Speaker 8: going to be five hundred seven three seven Maxes and
[10:33] Speaker 8: a large number of white body planes, and that would
[10:36] Speaker 8: be significant because the Chinese airlines haven't really ordered any
[10:41] Speaker 8: Boeing planes in a while and.
[10:43] Speaker 4: Just because they're mad at us.
[10:44] Speaker 8: They were mad at and there was COVID in the
[10:47] Speaker 8: middle of it. And so the Chinese airlines actually need
[10:50] Speaker 8: those planes because they don't really have a supply of Boeing.
[10:54] Speaker 8: I mean, they operate both Airbus and Boeing planes, and
[10:56] Speaker 8: they don't really have a supply of planes beyond the
[10:59] Speaker 8: twenty thirties. And so if they want to keep expanding,
[11:02] Speaker 8: if they want to keep adding more service, they need
[11:04] Speaker 8: those planes, and disorder is a way to get that.
[11:08] Speaker 5: Okay, so this order is a way to get that.
[11:10] Speaker 5: I guess if you're.
[11:13] Speaker 2: An airliner, you're thinking, what's the next thing Boeing is
[11:15] Speaker 2: selling that I want to get my hands on. I mean,
[11:17] Speaker 2: you know, not tomorrow, not next month, but down the
[11:20] Speaker 2: road as I plan my fleet.
[11:22] Speaker 5: Does boe have an answer to that?
[11:23] Speaker 2: Because it feels like the company has spent so much
[11:25] Speaker 2: time writing itself that it hasn't really put as much,
[11:29] Speaker 2: at least outward energy into innovating the next jet.
[11:33] Speaker 8: So that's that's the really big question that the CEO,
[11:35] Speaker 8: Kelly Ottberg's really looking at. So while obviously you have
[11:39] Speaker 8: the short term and the immediate issues that they have
[11:42] Speaker 8: to ramp up production, they have to ramp up cash
[11:44] Speaker 8: like cash generation reduce that debt pile.
[11:47] Speaker 10: The bigger question for Boeing is where do they go
[11:49] Speaker 10: from here?
[11:50] Speaker 8: And the seventry seven is basically a long derivative of
[11:55] Speaker 8: a nineteen sixties jet, and so for Boeing they need
[11:59] Speaker 8: to decide what the do next, and that will be
[12:01] Speaker 8: crucial in the sort of deeoperly that they have with Airbus,
[12:04] Speaker 8: and Airbus has been talking about how they are working
[12:07] Speaker 8: on their next generation of product that they're aiming to
[12:10] Speaker 8: get out in the middle of next decade, and that
[12:13] Speaker 8: would basically allow airlines to get something that's more fuel
[12:16] Speaker 8: efficient and allows them to fly greater distances with more
[12:20] Speaker 8: passengers on them. And so what we understand is that
[12:24] Speaker 8: Boeing is working on something that's more evolutionary rather than revolutionary,
[12:29] Speaker 8: because the last thing airlines need is uncertainty. And when
[12:33] Speaker 8: you have a product that's revolutionary, it always comes with
[12:37] Speaker 8: deathing issues. It comes with all sorts of complications with
[12:40] Speaker 8: new technology and how that whole measures together with existing infrastructure.
[12:45] Speaker 5: Yeah, we've seen that play UPFT exactly.
[12:47] Speaker 8: And so for the airlines, they want something that stable
[12:49] Speaker 8: allows them to sort of keep going as usual and
[12:52] Speaker 8: not sort of rocking the boat too much.
[12:53] Speaker 10: And that's really key for Boeing.
[12:55] Speaker 8: So how do you derive more fuel efficiency using evolutionary technology?
[13:02] Speaker 8: And that's where we see Boeing sort of going in
[13:05] Speaker 8: terms of their next products.
[13:08] Speaker 3: Ten seconds, if I get a big order today from China,
[13:10] Speaker 3: they can't even make the planes bowing.
[13:12] Speaker 4: I mean they have to build another factory or two.
[13:14] Speaker 8: So Boeing is in the process of opening a new
[13:17] Speaker 8: factory for the seven three seven, and they are looking
[13:20] Speaker 8: at ramping up production and they've been talking about how
[13:22] Speaker 8: they've got these short term goals. So they're currently looking
[13:26] Speaker 8: at forty two a month. They're going to ramp up
[13:28] Speaker 8: to forty seven a month, then fifty two a month
[13:30] Speaker 8: and sixty So essentially for Boeing that they have a
[13:34] Speaker 8: sort of stepped up plan to ramp up production depending
[13:37] Speaker 8: on how the FA sort of signs.
[13:38] Speaker 5: Off on it, and product will be in the US.
[13:41] Speaker 10: They will be in the US, so Boying produces all
[13:42] Speaker 10: its planes in the US. Stay with us.
[13:45] Speaker 4: More from Bloomberg Intelligence coming up after this.
[13:52] Speaker 1: You're listening to the Bloomberg Intelligence podcast. Catch us Live
[13:55] Speaker 1: weekdays at ten am Eastern on Apple Corplay and Android
[13:58] Speaker 1: Auto with the bloom Work Business Up listen on demand
[14:01] Speaker 1: wherever you get your podcasts, or watch us live on
[14:05] Speaker 1: YouTube AI.
[14:06] Speaker 3: It's impacting seemingly all parts of our life, all parts
[14:10] Speaker 3: of business and the economy, and that includes weather forecasting.
[14:14] Speaker 3: But maybe some shortcomings there. Our next guest has some
[14:17] Speaker 3: thoughts there at Caatham Kunda. He's a lecturer at the
[14:19] Speaker 3: Yale School of Management and he is a Bloomberg Opinion contributor.
[14:23] Speaker 3: You say AI isn't built for bad weather's Black Swan era?
[14:27] Speaker 4: What do you mean by that? Gotham?
[14:29] Speaker 10: Hey, Paul, good to be back.
[14:30] Speaker 9: So it's about weather forecasting, but it's not just about
[14:33] Speaker 9: weather forecasting. It's about the fact that we're using AI
[14:37] Speaker 9: all over the place as a tool to use the
[14:40] Speaker 9: past to predict the future, which is maybe the most
[14:43] Speaker 9: fundamental of all human activities, and it's incredibly important. And
[14:47] Speaker 9: it turns out that if as long as you have
[14:49] Speaker 9: enough and rich enough data about the past, AI can
[14:53] Speaker 9: be incredibly good at it, even better than the tools
[14:55] Speaker 9: we've used in the past, which in the case of
[14:57] Speaker 9: weather forecasting, involve planes and physics models and supercomputers and
[15:02] Speaker 9: weather satellites. So all of that is great. The problem
[15:06] Speaker 9: is that the weather is changing, so the weather of
[15:10] Speaker 9: the past may not be a good guide to the
[15:12] Speaker 9: weather of the future. And AI is incredibly good at
[15:16] Speaker 9: understanding stuff that happened within its training data set, and
[15:19] Speaker 9: incredibly bad about understanding stuff that's happening outside its training
[15:23] Speaker 9: data set. And as I said, that is about weather forecasting,
[15:26] Speaker 9: but it's not just about weather forecasting. It's about anywhere
[15:29] Speaker 9: where we're facing this problem of using the past to
[15:31] Speaker 9: predict the future. But the past is not like the future.
[15:35] Speaker 2: And in your column you write it will work right
[15:38] Speaker 2: up until it doesn't. AI models have a tendency to
[15:40] Speaker 2: fail in the biggest moments. And one example you give
[15:44] Speaker 2: that I think anyone who's over the age of eighteen
[15:47] Speaker 2: can relate to is the value risk models that banks
[15:50] Speaker 2: use during the Great Financial Crisis, where they relied on
[15:54] Speaker 2: past data to predict how much risk they can each
[15:57] Speaker 2: take on.
[15:58] Speaker 9: That's right, And that's another case of a tool that
[16:00] Speaker 9: works really, really well under normal circumstances, so well that
[16:04] Speaker 9: it gives you confidence, right, it makes you start to
[16:06] Speaker 9: rely on the tool. You might even enshrine it in
[16:09] Speaker 9: regulations so that when it fails, it's not just a
[16:12] Speaker 9: normal failure, it's a catastrophic failure.
[16:14] Speaker 10: So that's one layer of problem.
[16:16] Speaker 9: The second layer is that often these tools, it can
[16:18] Speaker 9: start to get used by people who aren't they people
[16:20] Speaker 9: who built them, and so who don't necessarily understand them
[16:22] Speaker 9: very well. I was at MIT during the to the
[16:25] Speaker 9: financial crisis, and I remember a very senior risk management
[16:28] Speaker 9: person at one of the banks at MIT saying to
[16:31] Speaker 9: a group of us, you know, I saw six sigma
[16:34] Speaker 9: events on three consecutive days, so the markets are weird.
[16:38] Speaker 9: And I just stared at him and I said, that
[16:40] Speaker 9: doesn't mean that the markets are weird. It means that
[16:43] Speaker 9: your tool has failed, because six sigma events three days
[16:47] Speaker 9: in a row will not happen in the entire lifespan
[16:51] Speaker 9: of the universe. Right, That's how rare we're.
[16:54] Speaker 10: Talking about is.
[16:55] Speaker 9: And this was a person who, you know, they were
[16:57] Speaker 9: very smart, but they didn't understand what this meant.
[17:01] Speaker 3: So, you know, I think one of the AI question
[17:04] Speaker 3: marks Gotham right now, is is AI going to create
[17:08] Speaker 3: jobs in this economy or destroy jobs in this economy?
[17:11] Speaker 3: I'm not sure we know that, do you have any
[17:13] Speaker 3: thoughts on that.
[17:14] Speaker 9: Well, I think we know that the answer is both,
[17:17] Speaker 9: it's going to create it's going to destroy some jobs,
[17:20] Speaker 9: it's going to create others. But the data on AI
[17:23] Speaker 9: job described destruction, the actual empirical data is pretty weak.
[17:27] Speaker 10: It looks more.
[17:28] Speaker 9: Like companies that over hired or that are looking to
[17:30] Speaker 9: cut costs and are kind of blaming AI for what
[17:33] Speaker 9: they did. But at the same time, there are going
[17:37] Speaker 9: to be some big transitions, and you know, all of
[17:39] Speaker 9: the standard models that we have, they're all about like
[17:42] Speaker 9: you know, bank tellers or things like that that tell
[17:44] Speaker 9: us new technologies don't create jobs, don't don't destroy jobs,
[17:48] Speaker 9: they don't destroy jobs and aggregate, but they destroy some jobs,
[17:50] Speaker 9: and the transition can be really hard for the people
[17:53] Speaker 9: who are undergoing through it.
[17:55] Speaker 2: So you are lecturer at the Yale School of Management,
[17:58] Speaker 2: How does education prepare young people? Graduates, those who have
[18:04] Speaker 2: some working experience, are going back to school and want
[18:07] Speaker 2: to come out, you know, with a better skill set
[18:09] Speaker 2: to prepare for this world.
[18:10] Speaker 9: So I think there are two things we need to
[18:12] Speaker 9: think about really hard here. One is that one of
[18:14] Speaker 9: the really interesting things about AI, at least as it's
[18:17] Speaker 9: currently structured, is it seems to be a magnifier, not
[18:20] Speaker 9: a compressor. Right, So if you think about like power tools,
[18:24] Speaker 9: power tools and construction equipment made brute physical strength less
[18:28] Speaker 9: important if you were working in manual labor. So they
[18:31] Speaker 9: compressed the differential and productivity between people who were physically
[18:35] Speaker 9: very strong and people who were less strong. But there
[18:38] Speaker 9: are other sorts of technologies that they increase the differences
[18:41] Speaker 9: in capability, so someone who normally would be twice as
[18:43] Speaker 9: capable becomes ten times as capable. AI looks more like
[18:47] Speaker 9: something in the second pool, right. It looks like it
[18:50] Speaker 9: takes people who are good and it makes them great,
[18:52] Speaker 9: but it's not that useful if you're not already an
[18:54] Speaker 9: expert in the field, and so someone has to create
[18:58] Speaker 9: those experts. That's the second That's what the one thing
[19:00] Speaker 9: that schools can do. The second thing is that schools
[19:03] Speaker 9: are you know, like at their best, and not all
[19:05] Speaker 9: schools do this all the time. I'm not defending that
[19:08] Speaker 9: saying that at all, but at their best, schools can
[19:10] Speaker 9: learn teach people to think independently of these skill sets.
[19:14] Speaker 9: And you're even seeing AI companies coming out and saying
[19:16] Speaker 9: that like humanity skills are really really important, that's what.
[19:21] Speaker 10: Yeah.
[19:21] Speaker 9: So, so something I've been I've been quoting at all
[19:23] Speaker 9: my programmer friends with great joy is when there was
[19:26] Speaker 9: all this turmoil before, people were saying they would you
[19:29] Speaker 9: know the responsibilities, but you guys need to learn to code, right,
[19:31] Speaker 9: And so now the AI is doing the coding, and
[19:33] Speaker 9: I said, you guys need to learn to ode.
[19:38] Speaker 3: I mean that kind of goes to one of the
[19:40] Speaker 3: things I'm learning from the young folks is it's the
[19:44] Speaker 3: ability to ask AI the right question, the prompts, yes,
[19:48] Speaker 3: the prompts, and in your world and your day to
[19:51] Speaker 3: day at a university environment, gotam, How's AI impacting your world?
[19:56] Speaker 9: So a lot in multiple different ways. So it's but
[19:59] Speaker 9: I'll it's not just learn the prompts, it's you have
[20:02] Speaker 9: to know enough to know when the AI is wrong.
[20:06] Speaker 9: That is actually I found the single most important skill.
[20:08] Speaker 9: So I used AI to help me create the syllabus
[20:11] Speaker 9: for a course.
[20:12] Speaker 10: It was amazing.
[20:13] Speaker 9: It took what would have been a week of work
[20:15] Speaker 9: and let me do it in a day. I was
[20:17] Speaker 9: blown away. But several of the things that it gave
[20:20] Speaker 9: me were wrong in ways that you had to be
[20:22] Speaker 9: an expert to understand. We're wrong, but another expert we've
[20:26] Speaker 9: seen this and thought I was need it right. So
[20:30] Speaker 9: that was a really important distinction. And what you're seeing
[20:34] Speaker 9: first is, you know, look, cheating is a huge problem.
[20:36] Speaker 9: Like we've had to crack down on that, like written exams,
[20:41] Speaker 9: take home exams. I think, like, I'm really sorry to
[20:44] Speaker 9: hear that, see that happening, But it's really happening. And
[20:47] Speaker 9: I tell my students on the first day, I'm like,
[20:50] Speaker 9: you probably can cheat. And you know, look, honestly, I
[20:53] Speaker 9: can't tell you if I'm going to catch you or not, right,
[20:55] Speaker 9: Like the tools to catch you are not that great,
[20:57] Speaker 9: but you're gonna cheat yourself. Like what you want here
[21:01] Speaker 9: isn't the grade. Nobody cares about your grades. What you
[21:04] Speaker 9: want is the learning, and you get the learning from
[21:06] Speaker 9: doing the work, not by asking trat ept to do
[21:09] Speaker 9: it for you.
[21:12] Speaker 1: This is the Bloomberg Intelligence Podcast, available on Apple, Spotify,
[21:17] Speaker 1: and anywhere else you get your podcasts. Listen live each
[21:20] Speaker 1: weekday ten am to noon Eastern on Bloomberg dot com,
[21:24] Speaker 1: the iHeartRadio app, tune In, and the Bloomberg Business app.
[21:28] Speaker 1: You can also watch us live every weekday on YouTube
[21:31] Speaker 1: and always on the Bloomberg terminal

Shows are listed from Apple Podcasts' charts and directory, and Podcast Index where it is used. Every show and episode belongs to its publisher and plays from the publisher's own feed; this site hosts none of them.