Linda Yao, chief working officer and head of technique for Lenovo’s Technique, Options, and Companies Group and vice chairman of hybrid cloud and AI options, joins the Me, Myself, and AI podcast to elucidate the group’s transition from know-how product firm to managed providers supplier. It’s now serving to organizations with the change administration required to implement AI within the enterprise.

She shares each a framework round pace, ease, and experience to facilitate this adoption, in addition to the 4 pillars of AI readiness that Lenovo guides its purchasers to attain. Tune in to this episode, additionally, for Linda’s perspective on the function of human connection in what she calls the period of inference, a time once we ought to deal with the implementation of maturing AI instruments.

Subscribe to Me, Myself, and AI on Apple Podcasts or Spotify.

Transcript

Shervin Khodabandeh: Keep tuned after at this time’s episode to listen to Sam and I break down the important thing factors made by our visitor.

What are the 4 areas that corporations have to prioritize for AI readiness? Discover out on at this time’s episode.

Linda Yao: I’m Linda Yao from Lenovo, and also you’re listening to Me, Myself, and AI.

Sam Ransbotham: Welcome to Me, Myself, and AI, a podcast on synthetic intelligence in enterprise. Every episode, we introduce you to somebody innovating with AI. I’m Sam Ransbotham, professor of analytics at Boston School. I’m additionally the AI and enterprise technique visitor editor at MIT Sloan Administration Overview.

Shervin Khodabandeh: And I’m Shervin Khodabandeh, senior accomplice with BCG and one of many leaders of our AI enterprise. Collectively, MIT SMR and BCG have been researching and publishing on AI since 2017, interviewing lots of of practitioners and surveying hundreds of corporations on what it takes to construct and to deploy and scale AI capabilities, and actually rework the best way organizations function.

Sam Ransbotham: Hey, everybody. Thanks for becoming a member of us. As we speak, Shervin and I are speaking with Linda Yao from Lenovo. At Lenovo, Linda is the vice chairman of AI options and providers in addition to the chief working officer and head of technique for [Lenovo’s] Options and Companies Group. Linda, thanks for becoming a member of us.

Linda Yao: Thanks a lot, Sam. It’s nice to be right here.

Sam Ransbotham: All proper, let’s begin with Lenovo. For our listeners who will not be conversant in Lenovo, are you able to give us a short overview?

Linda Yao: Sure. Lenovo is a world know-how firm. We function in 180 markets. We’re a part of the International Fortune 500. We began, many many years in the past, as a PC producer, after which … we acquired, in 2005, the IBM PC division after which determined to diversify our know-how enterprise.

We acquired the x86 information middle infrastructure enterprise from IBM. We acquired the Motorola telephones enterprise from Google after they’d purchased Motorola. And with that, we created what we name our end-to-end, pocket-to-cloud portfolio.

The most recent journey that we’ve been on the previous three to 5 years has been all about service-led transformation. Lenovo simply launched — three and a half years in the past — our Options and Companies Group, which is throughout transferring up the stack and answering the demand from our prospects to assist them, not simply with their {hardware} know-how wants but in addition with their software program, their providers and options, and producing these buyer outcomes.

Sam Ransbotham: So I’m guessing synthetic intelligence finally ends up being a giant a part of these options and providers. Inform us about that.

Linda Yao: Completely. I’m in all probability relationship myself right here, however this has been the third wave of AI to hit the enterprise since I’ve been working in know-how. The primary wave about 20 years in the past was when huge information actually started discovering itself in enterprise commercialized purposes, most notably high-frequency buying and selling.

Then we fast-forward 10 years to a few decade in the past, and that was across the time when AWS [Amazon Web Services] actually made computing extra accessible and democratized. And so we had a approach to make use of that huge information. At the moment, each enterprise was establishing its personal machine studying group. Each college was making an attempt to launch a knowledge science grasp’s program.

That was the place I actually noticed these kind of information science, machine studying, and AI purposes hit the workforce and use this know-how to make our workflows extra environment friendly and generate an increasing number of of these enterprise outcomes. But it surely was nonetheless fairly area of interest, so what actually excites me concerning the previous couple of years is the truth that generative AI [GenAI] has captured the creativeness. You realize, nearly all of AI purposes that we work on at this time will not be essentially generative, however what ChatGPT did a few years in the past was actually put AI on the fingertips of each shopper and actually put it into the psychological mannequin of each C-suite member, so it’s grow to be a little bit little bit of the rising tide that raises all of our boats with regards to know-how adoption. [It’s] tremendous thrilling. And that’s the reason at Lenovo, I’m very excited and proud to be launching the AI options observe.

Shervin Khodabandeh: Inform us extra about that. What does that observe do?

Linda Yao: We actually have embraced AI from an enterprise, business, and shopper perspective in a few alternative ways. The primary is by embedding AI into all of our current choices. I will provide you with an instance: We offer digital office options. We assist our prospects and their enterprises and organizations present help round finish consumer units. What we’ve been in a position to do now with generative AI is to hyper-personalize. For those who would have requested me 5 years in the past to tailor an organization’s IT help desk to assist particular worker profiles, even all the way down to the person consumer patterns on their gadget, may I’ve accomplished it? In all probability, however may I’ve accomplished it in a cheap approach? In all probability not. As a result of I might have needed to rent a bunch of analysts and a few supply hub someplace to actually analyze and undergo the info of each consumer’s patterns.

What we’ve launched now with this AI options observe is a set of providers — advisory, skilled, and managed providers — to assist corporations additionally undertake AI. That features the change administration providers to assist corporations actually assess whether or not their persons are prepared, whether or not they have the coaching in place, all the best way to know-how integration and implementation providers to make sure that you will have the appropriate GPU entry and configurations.

Sam Ransbotham: You’ve acquired a current examine that … is speaking about how difficult CIOs, specifically, are discovering dealing with this present wave of synthetic intelligence intention, I suppose. You identified that it’s introduced a number of consideration, however with that comes some difficulties for the C-suite. I believe one of many statistics that caught my eye, you mentioned one thing about 96% of CIOs anticipate rising funds, which doesn’t shock me, however 40-something p.c discovered that individuals have been struggling to see ROI from AI investments or didn’t count on it for 2 or three years.

Why are folks placing that cash in in the event that they’re not getting ROI?

Linda Yao: It’s an ideal query, Sam, and also you’re referring to our survey that we did for CIOs and IT choice makers throughout many markets, not restricted to any explicit geographic location. However what we discovered is comparable statistics throughout every of those markets.

Each CIO has some impetus to need to embrace this know-how and actually put it to work within the enterprise, however you’re completely appropriate. Fewer than half of them see fast ROI or see ROI on the horizon or see ROI that’s in a really achievable, sure approach.

We determine three essential gadgets that CIOs are on the lookout for. We name it pace, ease, and experience. Velocity signifies that we have now a set of choices that we name the AI Quick Begin. Inside one fiscal quarter, inside 90 days, whether or not it’s an enterprise AI drawback or an finish consumer AI drawback, we may help a buyer to spin up a proof of idea and a proof of worth in order that they’ll use this to display to their stakeholders why [they should] proceed the experiment or why to launch into manufacturing or why to deploy at scale.

From an ease element, we have now sense of what works and what works much less properly in an enterprise setting. And we have now curated a number of these greatest practices by area, by perform, into what we name an AI Library.

After which, lastly, is the experience. Loads of AI, if you get beneath the use case to really make it run and to make it run at scale, it’s all about GPU configurations. It’s all about information entry. It’s all concerning the engineering workflows. In order that degree of experience, that degree of depth, can also be what we’re in a position to contribute.

Sam Ransbotham: One of many issues that I believe is fascinating right here is that with all people paying a lot consideration to synthetic intelligence, one of many issues that your examine talked about was the chance of AI washing. I don’t suppose we’ve talked about AI washing a lot but on this present, and this looks like an ideal alternative to say that.

It looks like, with all people saying, “Oh yeah, AI is all the things,” there’s a bent to place a sticker on each answer and name it “new and improved, now with much more AI,” when, perhaps when you peel again the layers, there’s not fairly as a lot synthetic intelligence underneath there as you may count on. What’s the basis of all this AI washing? Is it an issue?

Linda Yao: You realize, Sam, AI washing is a little bit bit impressed by the phrase greenwashing the place a number of corporations may, you already know, with good intentions, attribute a number of actions to being extra environmentally sustainable or heading towards their ESG [environmental, social, and governance] objectives and perhaps get a little bit bit carried away or perhaps a little bit bit exaggerated. We discover the identical factor on AI washing, within the sense that everybody is so excited to embrace AI. And there’s fairly a degree of FOMO [fear of missing out] as properly.

You realize, these CIOs, these C-suite members, they don’t need to miss out on the pattern. We need to present that we’re making progress. We need to present that as AI and GenAI have hit the general public mindset that we as enterprises are additionally utilizing it to enhance our companies. And so, generally what occurs is that a number of issues and a number of options may get branded as AI, however, to your level, it would begin from a kernel of fact however won’t be the entire fact. It’s not essentially accomplished nefariously. I believe it’s simply because everybody’s so excited, however not everybody has come in control but on the most recent know-how, what it will possibly really do and what AI can’t do, and the way rapidly, or what it takes to implement it.

We have now discovered that the very best protection towards AI washing is absolutely simply to have a look at the details, to have a look at outcomes, as a result of outcomes don’t lie. And once we take a look at the outcomes that we’ve really achieved with AI, we’re in a position to actually dissect that answer all the way down to: What are the algorithms? The place did we entry the info? What’s the computational metric that we used to measure this ROI? And what forms of purposes and what forms of enterprise eventualities will we use it in?

I’ll provide you with a few examples of our personal early adopters of GenAI. Once I take a look at our personal C-suite, our chief advertising officer and our chief authorized officer have been two of the primary huge capabilities who noticed GenAI as a method to seize a number of the lower-hanging fruit when it comes to their automation wants. What we have been in a position to do for our chief advertising officer was to create an AI-powered content material technology platform that’s now in a position to create tailor-made content material for our … product brochures, buyer pitch books, function reality units — create all of that in a approach that has now saved them 90% on third-party company prices. That’s the kind of consequence that doesn’t lie. That’s how we all know that this isn’t an AI wash. That is really hitting the underside line.

Shervin Khodabandeh: I simply need to perhaps riff on this a little bit bit since you talked about the three unlocks to ROI: pace, ease, and experience. But it surely was additionally very a lot, I might say, tech-focused since you’re speaking to the CIO, I might think about. The half that I’ve seen a number of AI packages fail on the ROI aspect is … they’ll’t even get off the bottom with out these three issues that you just mentioned, proper? Since you’re going to construct stuff that’s not going to run or not going to run at scale, or the info goes to be far and wide, or they don’t have sufficient compute energy, so all of that’s only a utterly essential situation.

However then so many proofs of ideas simply type of sit on the shelf as a result of [of] the organizational side of the change that’s required and folks doing issues otherwise. You talked about the advertising instance for Lenovo, which was fairly profitable from what you’re describing. However there you’re inside the corporate and so that you’re attending to have an effect on how entrepreneurs and model and inventive persons are doing issues. How do you traverse the hole between the tech aspect and the precise utilization or as, you already know, Sam and I speak about, from the manufacturing to the consumption?

Linda Yao: I actually suppose that a big a part of it’s the folks aspect, to your level, along with the know-how aspect. Our methodology is absolutely centered on 4 pillars of readiness. So these 4 pillars of readiness are safety and information, folks, know-how, and processes. It sounds fairly easy on paper, however what we’ve seen is strictly your level, Shervin, that oftentimes, once we take a look at the Maslow’s hierarchy of wants, the primary hurdle that these stakeholders have to beat is throughout safety and information. They need to perceive, on the very primary degree, “How is AI and GenAI going to affect my safety posture? Is my information prepared? Do I’ve the appropriate guardrails up? Do I’ve the appropriate firewalls up?”

As soon as we get previous that time of what I might name danger administration, the subsequent factor they need to know is that if their persons are prepared. That’s why, as a part of our AI observe, change administration and AI adoption and coaching [are] an enormous side of it. Loads of what we do on this folks readiness evaluation can also be assist an organization perceive which of [its] departments are most able to embrace AI. Which job profiles, which job duties in that division may be extra prone to embrace AI, see the positive aspects, after which be a optimistic suggestions loop and a beacon for the remainder of the group?

We discover that one of many greatest obstacles to scaling AI, outdoors of know-how, is definitely the folks and understanding, for example, if the AI is supposed to enhance their duties or to switch a few of their duties. After which we have now know-how and processes, proper? Expertise, processes, and information [are] nonetheless by far the most important barrier for CIOs. [They are] nearly all of what they see with regards to inhibitors to scaling AI.

Sam Ransbotham: Once I hear these, all of them make sense to me and so they resonate. The concept of safety and folks and tech and processes make sense, however there’s one thing about them that doesn’t really feel very AI particular. That appears like the identical types of issues we’ve had with IT implementations for years and years. A part of that’s comforting as a result of we, in idea, ought to have expertise with this, however is there something completely different about synthetic intelligence with this course of that differs from the prior generations?

Linda Yao: It’s extremely astute, Sam, as a result of once we develop these 4 pillars of AI readiness, it’s really the identical areas and the identical classes we checked out for our digital transformation that has been occurring for many years and [is] not restricted to AI and definitely not restricted to GenAI. I believe what GenAI although has actually dropped at the forefront is extra deal with every of those pillars otherwise.

For instance, the safety query: It’s so far more prevalent now simply due to the sheer quantity of knowledge that AI requires and the quantity of knowledge that AI now generates. Understanding the place the info lies, the place it’s being accessed, how we govern it, what forms of obstacles we put across the utilization of that information and across the newly generated information that turns into hyper-important.

Shervin Khodabandeh: Linda, you’ve been doing this for a while, so you will have some longitudinal perspective right here as properly. Do you see within the information because the creation of generative AI that CIOs and corporations, extra broadly, have a greater sense of the place they’re going with AI or a worse sense of the place they have been going with AI?

As a result of Sam and I did some analysis right here. The rationale I’m asking is there was a time the place [we] would ask hundreds of parents that we surveyed, “Do you will have an AI technique?” And again in 2017 and 2018, 20%, 30%, 40% would say, “Sure, however we’re engaged on it.” By 2018, 2019, 2020, most individuals would say, “Sure, we have now a technique.” And by 2021, 2022, most corporations would have an AI technique, and so they have been investing in [it] and so they have been getting some outcomes. However [in] 2023, 2024, what we noticed within the information was a smaller proportion would say, “Sure, we actually, actually know what we’re doing with AI.”

And I’m questioning, how does it have an effect on the CIO’s outlook from the place you’re sitting sooner or later, when it comes to how properly articulated the imaginative and prescient and the plan and the path [are]?

Linda Yao: That’s such an fascinating query. AI technique is a subject the place the extent of precision and certainty that we’d see in another methods of know-how adoption is just not essentially there. However I don’t see that as a foul factor. Earlier than, I discussed the couple of waves of AI adoption within the enterprise that I’ve skilled even in my profession. Earlier than, they have been extra focused, proper? Earlier than, it was not each division that had the power to generate and curate a lot information to be the gasoline for AI.

Earlier than, not each group had the ability set essential to be growing information science algorithms, to be implementing machine studying engineering purposes, and to be taking these issues to scale. It truly is at our fingertips. The obstacles to entry for each finish consumer, for each shopper, have actually been lowered — when it comes to the forms of AI we will experiment with, the forms of AI and information that we will create ourselves, after which, subsequently, how rapidly it will possibly proliferate. However I do suppose that one enormous weapon within the arsenal of those C-suite members is that AI is now so accessible, it has captured the creativeness. We will every now grow to be an increasing number of educated on AI ourselves.

Shervin Khodabandeh: Linda, we have now a section right here that we name 5 questions. I’m going to ask you 5 questions, and also you inform me the very first thing that involves your thoughts.

Linda Yao: Phrase affiliation. Deliver it on.

Shervin Khodabandeh: What do you see as the most important alternative with AI proper now?

Linda Yao: I believe that one of many greatest alternatives I see with AI proper now’s how we at the moment are transferring from the period of coaching to the period of inference, proper? What I imply by that’s we have now spent a number of time and a number of funding now honing the instruments, honing the muse fashions, honing the platforms to actually harness all this energy of synthetic intelligence. Now, I believe we’re going to see an actual renaissance and resurgence round how will we really put that to make use of? What are the purposes going to be? What are the use circumstances going to be? And the way is that going to make an actual distinction for you and me?

Sam Ransbotham: That’s good.

Shervin Khodabandeh: What’s the greatest false impression with AI?

Linda Yao: I believe one of many greatest misconceptions is, once more, this trepidation about AI coming for our jobs, AI coming for our lives, AI coming for our consciousnesses. Perhaps I’m too optimistic, and I prefer to consider in a world the place we can not simply coexist however really wield AI to make ourselves higher and to make society and business higher. To me, that’s a giant false impression, that one way or the other we as people will lack the power and one way or the other be overcome by this AI that we create. No, I believe we simply must go about it responsibly and intelligently.

Shervin Khodabandeh: What was the primary profession you needed? What did you need to be if you grew up?

Linda Yao: I needed to be an airplane pilot. I needed to fly fighter jets, and that’s really one of many causes that led me to my earlier employer.

Sam Ransbotham: You have been at Boeing earlier than.

Linda Yao: I used to be at Boeing earlier than, and I used to be in a position to work on the flight line for the F-18s, and it was one of many highlights of my profession there.

Shervin Khodabandeh: Fantastic. When is there an excessive amount of AI?

Linda Yao: When is there an excessive amount of AI? You realize, there are a number of purposes now that actually implement the human within the loop. And I believe the human within the loop is a approach to usher in one other guardrail that actually focuses on protecting AI throughout the bounds of the place it’s meant to be. I don’t need to make that sound too sinister, proper? Like, the AI is a naughty creature that’s all the time making an attempt to get out, however there are undoubtedly use circumstances the place AI is just not the relevant know-how, proper? In use circumstances the place we don’t have the appropriate information or sufficient information or an unbiased view of the info, these are the use circumstances the place AI is just not going to work properly as a result of information is the gasoline for AI. So we’ve recognized for ages now, in any kind of know-how, rubbish in, rubbish out. And that’s the place we have now to actually watch out.

Shervin Khodabandeh: What’s the one factor you would like AI may do proper now that it will possibly’t?

Linda Yao: You realize, for AI — I believe it’s really coming quickly — I actually want that AI may grow to be much more hyper-personalized [and] actually assist every of us as people, not simply make ourselves extra productive at work however actually make ourselves extra, you already know, fulsome and extra engaged in our lives outdoors of labor, with our households. Closing these connection loops, actually closing the gap and a few of these obstacles.

I believe that aspect of AI that helps us grow to be higher — I’ll say, emotional-sensing perceptive folks — might be coming, to be trustworthy. I see the extent of hyper-personalization. I see the developments we’re making in AI feeling and sensing and perceiving extra round us. I believe it is going to come, however getting previous the AI that’s throughout chilly, laborious details and outcomes, and attending to one thing that is a bit more holistic, that’s one thing I want AI may do now.

Sam Ransbotham: Linda, thanks for becoming a member of us. That appears like a good way to finish — eager about how hyper-personalization may help us make connections with folks and never with know-how. Thanks for taking the time to hitch us.

Linda Yao: Thanks for having me, Sam and Shervin.

Sam Ransbotham: We simply completed speaking with Linda, and a pair fascinating factors got here out: the emphasis on the ROI and the problem, and perhaps the short-term horizon for ROI, but in addition the dialog went in a little bit completely different path. Whereas we began off speaking about organizations and know-how, we ended up speaking about folks and human connections. Each of these are form of fascinating.

Shervin Khodabandeh: It’s fascinating Linda’s commenting on that, too, as a result of it’s a indisputable fact that earlier than generative AI, corporations have been spending rather a lot on AI, after which 2023 and 2024 got here, after which they doubled down [on] these investments. They acquired a move in 2023 and 2024 when it comes to displaying returns, notably a few of … this AI washing factor that she was speaking about, which is like, “Oh, no, it’s essential. We’ve acquired to do it.” This notion of AI washing is much like greenwashing, so [it] completely is sensible that investments went in [and some achieved] nice returns. We see a combined bag, proper?

I believe that’s additionally an actual factor that they’re seeing from their lens. Sam, you identified that this looks like a digital transformation form of a factor, so what’s completely different? I believe that it’s acquired to be completely different as a result of the endurance isn’t there. The speed of change isn’t the identical. And I believe a number of it’s within the suggestions loops and the educational loops and making an attempt sooner and studying sooner and altering the character of labor. That was the primary factor for me.

Sam Ransbotham: I believe if I used to be sitting in a CIO suite and such as you mentioned, “Oh, you get a move for 2023, 2024.” Nicely, what number of extra passes will we get? I really feel like in some sense we’re saying, “Oh, no, this time give us a few years. This time will repay.” What number of instances can we attain out and make that promise?

Shervin Khodabandeh: Precisely. I believe that is going to be a troublesome 12 months: 2025 and 2026 are going to be powerful years for corporations, and they should present affect.

The opposite factor I actually preferred, which was additionally how we ended the present, was on this notion of AI as a approach for us to be extra human, which is a little bit little bit of doing a little bit little bit of jujitsu on AI itself, proper?

I believe she’s proper. It’s not that far off with all the things that multimodal can do with tone and language and context and all that. It may be a coach. I imply, technically, it may be a coach and assist us perceive ourselves higher, be extra conscious, perceive the impacts that we have now with our phrases and on different folks. Minimally, it may remind us to select up the cellphone and name the parents we haven’t known as. It’s fascinating.

Sam Ransbotham: That appears like AI washing. That doesn’t should be AI.

Shervin Khodabandeh: What do you consider that, although, Sam?

Sam Ransbotham: I imply, a part of me wonders generally if we’re looking out so laborious for there to be a human connection that we grasp at straws. However alternatively, you’re proper, that does really feel like a possible use, that there’s no cause to consider that AI use can solely have an effect on this a part of our lives and never different elements of our lives. It will be type of naive to suppose that it could be restricted in scope to only our office effectivity. I’ve some hope for that.

Shervin Khodabandeh: However I may see it will possibly go in scary methods.

Sam Ransbotham: I do fear that it begins to really feel [inauthentic] and [disingenuous] that, you already know, if I’m being good to you, Shervin, however solely as a result of I’ve acquired a … pink gentle displaying in my monitor that my tone wants to enhance after I’m speaking to you, you’re going to see proper via that. You’re going to know that I’m solely doing it to move my content material filter. That doesn’t really feel genuine. That looks like the know-how is placing [up] a barrier.

Shervin Khodabandeh: Precisely.

Thanks for listening, everybody. Subsequent time, Sam and I communicate with Chandra Kapireddy from Truist. Please be part of us.

Allison Ryder: Thanks for listening to Me, Myself, and AI. Our present is ready to proceed, largely, resulting from listener help. Your streams and downloads make a giant distinction. When you have a second, please take into account leaving us an Apple Podcasts evaluation or a ranking on Spotify. And share our present with others you suppose may discover it fascinating and useful.