Machine transcription with reviewed terminology and names. Some recognition errors may remain. Timestamps refer to the livestream; subtitle downloads are timed to the session clips.
6:02:05Good afternoon, everyone present here.
6:02:09It's been a great day and an absolute pleasure to be here.
6:02:15The topic of my talk today is everyone has AI, who has the advantage?
6:02:24I'll be talking about what the orgs that are doing, that are leading in AI with their implementation and adoption,
6:02:33what are they doing differently?
6:02:37so before that, who am I and what's my POV?
6:02:41So I'm Dr. Sid Pednekar. I am an AI technology and data exec in asset management.
6:02:50I've spent most of my tenure in hedge funds and investment in academic research.
6:02:57What I've done and focused on is lead strategy and build specialized teams towards three main verticals,
6:03:06optimizing efficiency, driving PNL, and reducing costs.
6:03:12So I'll be following the discussion along with two hypothetical companies,
6:03:20Company A and Company B, right at the slider below.
6:03:24These are hypothetical, you know, with made-up workflows, but we'll be following their journey along the rest of this presentation.
6:03:36So, everyone has AI. Who has the advantage?
6:03:41Before we get started there, first let's see what the lay of the land is.
6:03:46And recent studies have shown that only about 37% of enterprises that were surveyed are
6:03:55showing a positive contribution to their EBIT.
6:03:59And only about 6% of those would be qualified as high performers.
6:04:05So that's about 5% contribution to their EBIT.
6:04:08Other studies earlier in the year have shown that about 66% are saying they're having a productivity increase,
6:04:17but only about 40% and 20% are actually seeing cost reductions and revenue growth.
6:04:24So as we know, access of AI is widespread, yet the impact remains uneven.
6:04:35So, the question I would ask, or how I think about this, is really in terms of three verticals.
6:04:45And the way I think about these three verticals should apply to organizations and individuals
6:04:52alike, and it should apply to functions, regardless of where you are in that organization.
6:05:01So the three main vertical pillars that I think of are, number one, design the right problem.
6:05:08Number two, build shared capabilities.
6:05:12And number three, compound the advantage.
6:05:16Now before we dive into each of these pillars, there is an idea that I think goes through all these verticals.
6:05:24And I think the best leading firms are taking advantage of that and capitalizing on it.
6:05:31As we know, throughout time, the only constant in this world has been change.
6:05:38What's different in the AI and superintelligence era is that intelligence has become cheap.
6:05:45Regardless of what your unit of intelligence is, whether it be software, analytics, recommendations, reports,
6:05:54AI is creating an abundance in all of it.
6:05:56And as we know, more answers don't necessarily create more value.
6:06:03So what is scarce?
6:06:06What is the scarce resource in an abundant world?
6:06:11What are companies taking advantage of, the leaders at least?
6:06:15It's judgment.
6:06:18What is judgment?
6:06:20Judgment is choosing what matters and how to act
6:06:24when there's no single obvious answer.
6:06:26It's the creativity that lets you explore what possibilities are worth exploring.
6:06:31It's the taste that tells you what good looks like in your given context.
6:06:36It's the decisions where you can weigh what trade-offs are worth making.
6:06:44is ubiquitous.
6:06:45It exists throughout your organization on a daily basis.
6:06:49Your C-suite, the C-suite is asking questions, taking advantage of judgment all the time.
6:06:55So are the software engineers, the AI leaders, the portfolio managers, and the data scientists.
6:07:01So the question that they're focused on is, given judgment is a finite asset, a limited capacity,
6:07:13the main question becomes, where is our best judgment worth spending?
6:07:18I'd like for you to hold on to that thought.
6:07:21With that, we'll dive into our first pillar.
6:07:23Design the right problem.
6:07:25So questions, a question that the leading organizations are asking all the time is they're obsessed over what is the right problem to solve, right?
6:07:39It's tempting to start with a tool.
6:07:41Today we're surrounded with better models, harnesses.
6:07:45We're surrounded with all these AI tools.
6:07:47We're surrounded with co-pilots.
6:07:49But that isn't the way about, is it?
6:07:52At least that's not the way the leaders are approaching it.
6:07:55Now, let's go back to our hypothetical two companies.
6:07:59Let's say they're from asset management.
6:08:00Let's say we're tracking the workflow of two research analysts
6:08:05in these hypothetical organizations.
6:08:09Well, let's say organization B starts with the tool, the co-pilot.
6:08:14They're asking, where can we add AI to the analyst's work?
6:08:19They're looking at the current existing process to say,
6:08:22earning transcript comes in, I discovered a summarization workflow, let me pass all of
6:08:28this in through my fire hose, analysts get earning summaries faster, win-win, right?
6:08:36Well, not so much, at least not compared to the guys who are leading.
6:08:41And such a workflow of being a faster horse introduces one key nuance.
6:08:47That is, you are losing context as you have information, I guess, reduction as part of your summarization.
6:08:55So what is a hypothetical organization A doing in this?
6:09:00Well, they're obsessed on the right problem to ask.
6:09:02They could be asking, why are our material signals reaching us too late?
6:09:07Where do we focus the attention, the AI, the bandwidth, the capability on?
6:09:12Where is our bottleneck, right?
6:09:14And so it becomes this redesigning of the workflow,
6:09:18or at least a re-evaluation on what to focus on.
6:09:21That's what they're doing different.
6:09:23Now, this consideration isn't particular to finance.
6:09:26It's across the realm and function.
6:09:29You're asking the same question on what to focus on,
6:09:33whether it be research capacity, decision speed,
6:09:37quality of your data, latency, reliability.
6:09:41It traverses all of that.
6:09:44studies show this.
6:09:46They show leading studies, on average,
6:09:50spend their attention on fewer higher quality decisions
6:09:54than the other cohort.
6:09:56So leaders are being more selective
6:09:58in where they focus their attention
6:10:00and what they focus their attention.
6:10:02So it's great.
6:10:03The leading companies have an idea
6:10:06of what the right problems are.
6:10:09Where do they go from there?
6:10:10Well, they also concentrate their investments in the highest value workflows.
6:10:18I get it, and I've been culpable of something similar,
6:10:22where AI is almost creating too many opportunities.
6:10:27And experimentation can be a good thing.
6:10:30But one needs to separate experimentation and the resulting innovation theater,
6:10:35or worse, the productivity theater, which actual value generation.
6:10:41Going back to our two hypothetical companies, what does this mean?
6:10:45Company B is looking at how do we launch the most number of AI pilots?
6:10:50And what that ends up looking like is a document summarization bot here,
6:10:55a co-pilot there, a data retrieval bot here.
6:10:59It's scattered all over the place.
6:11:01whereas what company A is focused on in their investments
6:11:05is potentially use cases like end-to-end workflows
6:11:10on the research pipeline.
6:11:13They're focused on aligning engineering, analytics, AI, capacity,
6:11:19and judgment towards a singular focus.
6:11:23And research has shown that the leaders spend more than 80%
6:11:29of their investments on these high-value use cases,
6:11:34transforming key business functions
6:11:36and creating new offerings.
6:11:39Perfect.
6:11:40We know they're focusing on finding the right problem,
6:11:44and they're also spending deep investment
6:11:48in select workflows.
6:11:49Well, they're also doing something different.
6:11:52They're reimagining how work is done.
6:11:56I think this is key.
6:11:59And it goes back to what we discussed previously, which is judgment.
6:12:03They're redefining and reorganizing functions, roles, capacities
6:12:08around this key element of judgment.
6:12:12What does that mean for our hypothetical company?
6:12:15Well, company B might be focused on, as we discussed,
6:12:18I'm getting my earnings transcript, I'm getting my research reports.
6:12:22How do I summarize it? Maybe I make that faster.
6:12:25Maybe I update my financial models faster.
6:12:27Maybe I send my recommendation to my PM faster.
6:12:31Company A is focused on an entirely different ballgame.
6:12:36Hypothetically, they could be asking the analyst to define the investment thesis,
6:12:40and they're watching and gathering evidence that could refute that thesis or support it.
6:12:46They'd be monitoring continuously, they'd be investigating exceptions,
6:12:52and they could be deep diving on select examples
6:12:59that could lead to actual generation of value.
6:13:03So the winners don't just automate, they redesign.
6:13:07studies show that the cohort of high performers
6:13:10are 75% more likely, or in practice,
6:13:13they actually do redesign their workflow
6:13:15versus 25% of the cohort that doesn't.
6:13:20So where does this leave us with our two hypothetical companies
6:13:23and designing the right problem? Company B adds AI to existing workflows, they're completing individual projects and they're moving to the next initiative, whereas Company A is rethinking how work is done. They're redesigning processes around AI and they're obsessing over the right problem. That's the difference in how organizations put AI to work.
6:13:46Great, so they've picked the right problem.
6:13:49What happens next?
6:13:51Well, we all know in order to actually enable that capacity,
6:13:55you need to build share capabilities across the org.
6:14:00And what's a trend that's seemingly emerging for some time now
6:14:05is that the lines between technology and business are converging.
6:14:10And we're seeing this everywhere.
6:14:11We're seeing engineers that are moving closer to business outcomes.
6:14:14We're seeing business professionals that are building tools and automations.
6:14:19And as you can see and imagine, you can have forward deployed engineers,
6:14:24you can have forward deployed product managers,
6:14:26but you could also have the inverse, which is forward deployed portfolio managers.
6:14:31And it's really this combination of getting the right people in the room
6:14:36by breaking down organization silos that the leaders are really leaning on.
6:14:43studies have shown that more and more of the large enterprises are viewing their technology arm as a business growth function
6:14:52and that's creating an edge for these companies
6:14:58so the other sort of vertical inside building shared capabilities that is probably a personal favorite of mine
6:15:06it's standardizing the core and personalizing the edges
6:15:11And this is something that isn't new to the technology world.
6:15:15We've built infrastructure.
6:15:16We've built shared modules and system design.
6:15:19But what's interesting here today is that's taking a very different shape with the capabilities out there
6:15:25and with the org and the breaking down of silos.
6:15:27And this could look very different depending on where you sit.
6:15:33You could be building shared cores
6:15:35for a team that looks very different
6:15:38with analysts and clients.
6:15:40You could be doing that with a bunch of analysts,
6:15:43or you could be doing that for software engineers
6:15:46or data scientists.
6:15:48With the data scientist example in particular,
6:15:51that could look very interesting
6:15:53where instead of having one-off workflows
6:15:57that is really common in the industry,
6:16:00you could build a shared core
6:16:02of the data sets, their semantic layers,
6:16:05the trade-offs of what worked, what didn't,
6:16:08what some data sets do good of,
6:16:10where are the gaps, and so on and so forth,
6:16:13such that each new personnel
6:16:15becomes almost a plug-and-play module
6:16:19in the system where they can then stand
6:16:21on the shoulder of their predecessors.
6:16:24So what should the next person not have to rebuild?
6:16:29now what's also interesting is um you know the enterprise wasn't really built for agents um agents
6:16:37they they work across workflows and applications they they take context they uh deliberate they
6:16:43take actions whereas really the enterprises are focused um previously on a single human
6:16:51or a collection of them interacting with a single application.
6:16:55So the leaders are cognizant that something needs to change
6:16:59and that they need to modernize their infrastructure.
6:17:02And they're very self-aware of where they are in their journey.
6:17:07And so studies show that 71% of the orgs
6:17:11are modernizing their core infrastructure to support AI.
6:17:16and 6% to 10% or 23% of them are investing 6% to 10% of their annual revenue
6:17:23in core system modernization to take advantage of this.
6:17:27And a common complaint here or a common challenge here is data quality
6:17:33and the ability to reuse data.
6:17:36So, great. We have the right people in the room.
6:17:40We have the right knowledge there.
6:17:45What do we do next?
6:17:47And then comes the concept of, hey, how do we make this knowledge reusable and accessible?
6:17:52And that's where we have, depending on the adoption and where they are, you can have different types of team agents,
6:17:59where you can have expert agents being subject matter experts.
6:18:03You can have workflow agents that work across streams like Slack bots that are creating and opening tickets.
6:18:09they're closing them, and what have you.
6:18:11And then you can have really chief of staff kind of agents
6:18:14that keep a tab on what the state of the organization is
6:18:18and what needs to be done.
6:18:20So, you know, going back to our workflows,
6:18:23A, again, is looking at this very differently
6:18:25in terms of shared capabilities.
6:18:27Now, the last pillar here that I want to draw your attention to is,
6:18:32well, when does this system really compound?
6:18:35What happens when it's deployed?
6:18:39Well, the leading companies, what they're doing is they're actually thinking about it slightly differently.
6:18:45What are failures that can give us an edge that we can learn from?
6:18:50If there's a success, how do we make it part of the system?
6:18:54And the way they think about this is, you know, the individual learned but did the system.
6:18:58Is there a way, perhaps, when they've encountered something where they need to make a correction,
6:19:03that that knowledge is fed back into the org,
6:19:06either through a new tool or context or what have you.
6:19:10And what's key here is preserving the evidence along with those learnings.
6:19:18So really the last vertical here is AI freeze capacity.
6:19:24Where does it go?
6:19:25Being cognizant that you need to redesign your system
6:19:29and really focus it on key metrics that you're evaluating,
6:19:35you're monitoring, and you're creating a flywheel off.
6:19:39So where does this leave us for companies A and B?
6:19:42You know, company A has their entire work redesigned.
6:19:46They're building capabilities at Compound,
6:19:49and they have shared capabilities built.
6:19:52So I want to leave you off with one final question.
6:19:55Everyone has AI. Who has the advantage?
6:19:59Thank you.
