Introduction

Good evening, everybody. Thank you for coming out in all these puddles to be here.

Speaker background

I just want to tell you a little bit about my background. My name is Samantha. It's nice to meet you.

I grew up in tech and ops, and I grew up in like big enterprise banks doing really boring stuff, but also decided to kind of change my life and worked in a lot of innovation labs and started productizing, you know, companies and parts of companies that weren't doing work that way.

I became kind of an agilist I really love like how people work and was trying to get to market things very quickly and and there were you know really famous jobs I had of translating between really smart tech folks which are in this audience today and old farty bankers like that would be kind of like one of my jobs and saying like hey this machine learning thing is not crazy like you can make better lending decisions based on this and I'll talk a little bit about this tonight because this is happening this translation and this like ability to do this is happening in a amplified scale right now and so I want to talk

tonight about how even if you're technical and you're making things or if you're a data scientist and you're creating these like beautiful outcomes where there's like a little bit of a disconnect so that's a little bit of my talk tonight I'm an executive coach now too so you'll see a lot of coaching talk

A quick AI mindset check-in

in here you'll see just a quick check -in I do this with pretty technical people every week in a weekly AI forum that I run but I doubt there's a one in here but are there any ones on this scale that are in this room that are kind of

petrified of AI and what it's gonna do believe it or not at universities you'll see hands raised here you got a one I got you I'm so proud of you for saying or one technical feudalism is what you're concerned about I'm with you

brother okay and then twos we have any a twos in here I was at a you know a team that was a digital team and the average in the room was a 2 .5 and it's a digital

team and a tech company by the way any threes in here all right very good got fours. It's good to know who I'm talking to. Got a four.

Fives. I'm expecting some hands here. Yep. And sixes. Yeah.

And remember, I want you to know that this is mindset. This isn't actually like your ability to vibe code. This is a mindset thing. And this is like a really big deal.

And there's no right or wrong answer here. This is just kind of like how we're kind of headed and the way people are thinking about work. So it's already cool to see like we're like an average

Why AI Isn't Scaling as Fast as the Hype

3 .7 in this room okay so what's happening right now we're throwing out a lot of AI making a lot of cool things but kind of companies are trying things but very little of it is getting scaled in like a massive amount we're going to talk about some of the tries that people have put up there to do that so we're

moving at like warp speed making cool stuff we're doing neat stuff at our desks they're doing things in tiny pods but decision ownership and and this has This has been talked about in each of the talks before me today.

From cool prototypes to owned workflows

What people are doing with this, how they're deciding. Take a little piece of workflow. Do not try to do too much. Like, who will own that?

Oh my gosh, what if we go in the wrong way? This is the kind of stuff that's important. So it's supposed to remove our repetitive work.

I had to make a summary slide for here, and I walked in the kitchen this morning. I'm like, oh my god, I'm going to make a summary slide. I'm like, oh my god, I can use chat GP to do that.

Yeah, it's supposed to take away some of our little thought work so we can do other things, like actually eat breakfast. and so we can focus on the higher value thinking I'm launching a business on

Monday so it was really great to not have to sit down for 30 minutes and go through this whole thing and decide what my summary statement was gonna be um but who thinks that this like whole thing feels harder than it was supposed to

feel raise of hands self feels harder right this was supposed to like save our lives and like be so easy and just even the little things if you're not writing a 50th birthday speech, this is hard. This is harder than it's supposed to be in kind of scaling

and those things. It feels a little bit confusing. It's confusing.

Why leadership teams feel stuck and confused

I do executive coaching, so I'm actually facilitating offsites for leadership, and we're having conversations about what they're going to do for 2026.

They are very confused about the things that they thought they were doing for 2026 and where AI sits in there. This is confusing times, and we're not sure where to put our money.

a lot of words don't worry but I put these up here to talk about like executives are trying to figure out like if AI so to help us like why is it like kind of not working or executions getting a little bit weird if you're a

technical person here you're making something beautiful and clean and nice like why isn't it people using it the way you want it let me just give you a perfect example not picking on any company but when you put a something

like Microsoft Co pilot in there have you taught people how to prompt like Like, an FYI, they're making a lot of money on that cloud space, right? So when you're not a good prompter and you're going in there and asking the wrong questions, every single time you're doing that, they're charging in the cloud, right?

So they don't want you to prompt that well, do they? I don't know. If you have them in your company, ask them to come in and do some office hours on prompting

so that you're using less cloud space. But this kind of thing is, like, very complicated right now. And if you're in data, you're seeing these outputs.

You're seeing this stuff. Why aren't people adopting this as quickly as you thought that they would when you make a beautiful dashboard for them any comments or questions or

examples I want this to be interactive any comments about this or experiences anybody's having you want to share okay all right keep having beers and then

The Paradox: Cheaper Intelligence, Higher Stakes

we'll get this more open so we're in this like paradox like the cost of intelligence and like kind of summarizing things is like brought down we've compressed these timelines and having to do things I can write a summary in 10 seconds and like check it but you know we have to kind of make

judgments about these things is that the right way I don't want my you know LinkedIn to look like it's not the deadlines it's the people it's why we're here you know you know that sent those sentences those three I and I'm a victim of this as well I've done it and now I've killed in my business partner we

have to stop so the other thing is is that when we amplify some of these these things the stakes are higher and I'm gonna show you some fun examples of

Decision design: the missing layer

that in a second so we haven't solved for this decision design which everybody who's technical in here really needs to think about where he said before Sahil was like making a point which was pick a small thing that somebody's very much an expert out and isolate that one thing so somebody knows whether that goes awry or

not and they know like what the drift is and like what good looks like we're not really doing that we're trying to take stuff like off our backlog make a thing and it's bananas okay so where where are things breaking this could be a fun

Where Things Break: Real-World Failure Modes

couple slides for you here's a bunch of organizations that forgot that the internet is biased so they like gave women like crappier credit limits and you know put people in jail based on their color and longer and decided they were riskier pretty shady important things were going on in here Zillow lost

like half a like million dollars like somehow with how they were calculating what's going on but the point is is if you have it if you're making something and it's cool you also need to think about the lifecycle of that who's involved and like who really needs to be a stakeholder in this product otherwise it's not going to scale so we can make a lot of really cool stuff I'm sure this

room could make an accelerator and we could like make an incredible product in like five days in some sort of hackathon but whether or not people are gonna be be able to use it's a big deal right now so a lot of words again these are just showing you know

Overload, trust gaps, and adoption resistance

one of the things that's happening is that you're coming at people first of all who feels like they're doing two jobs at work right now if you work in a corporate job entrepreneurs you're already doing 10 jobs but who feels you're you're at least doing two jobs entrepreneurs are doing 20 10 to 20 can you imagine if a bunch of bells and whistles and insights start pounding you out like every single day trying to tell you here's some updates here's some things here's some insight think about this for it's too much and this is happening in

cyber this is happening in healthcare it's confusing it's confusing for nurses and people that are in the front lines trying to do their jobs the way that they were it's taking a little of intuition out and so what's happening is is that people are starting to ignore it because it's just too much intelligence too much of the time and in enterprise environments you know it's great so we

have this insight and it's come to us but people are still slow people are still where does this information come from there are little resistance sometimes there's some trust that needs to be built up there's a lot that's going on there and they're not sure if they're gonna lose their jobs if they're gonna lose some sort of power if some priority they have is going to be taken away so there's some ego stuff going on as well which is also an inhibitor of

High-stakes mistakes and the need for humans-in-the-loop

this activity here's where you don't want to be on this extra extra page so I've got some great stories here my favorite story on here is publishing a summer reading list by two newspapers of books that are not real now this is where a real human in the loop opportunity some of these are really sad

and scary this one where they talked about the surgeries it's actually a very very cool technology that uses imagery and you kind of try to find the smartest, safest path to get to an organ inside a body. But they were misrepresenting what the organs were. In only like ten cases, it wasn't a lot, but they didn't know what an organ was, and so like bad things happened. So this is, you know, kind of frightening.

Like we need, you know, where does that decision lie that we need to own? And I know I'm not like kind of teaching you a totally new thing, but But I am trying to teach you this, because this is happening. And this is happening in surgeries, and it's happening out there. There is an entrepreneurial thing that told people to break the law, basically, in New York. And if they did that, they could have gotten in trouble for entrepreneurs. So there's a lot happening.

There's some very serious, I read this article from Reuters about nurses and really difficult, like one sensor's telling me something. And when I see these two pieces of information, this is what I do as a human being in a situation from all my training and then somebody else comes in and says when this and this happens and this is what the AI is kind of like telling us and this is what we do in a doctor had a third idea it's a really interesting article you should go check it out but

it's wild another favorite on this page is I know everybody already knows about this but the people that deleted the or the coding tool that deleted the the production database and then lied about it when they asked about it, that was super dope. Anyway, so this is kind of more of the philosophy thing.

AI as a Mirror of Organizational Health

AI is kind of a mirror.

And I was having a great talk with Sahil before this started as well, which was,

if you didn't organize your data in 2015 when you were supposed to and when Hadoop was running around in all these buildings and telling you to put stuff in data lakes the right way, you're in trouble now.

Like, you can't really use this that well, right? If your stuff is sitting in a bunch of ponds, you're kind of in trouble.

Data, agile prioritization, and the limits of command-and-control

1If you didn't learn how to do agile teams and truly prioritize and teach your teams how to, like, do the highest value, highest priority thing first and really have deep conversations about value, you're in trouble now. Because we need to start prioritizing, like, what we're doing. And if you have a language for that and a language for risk and stuff, you're going to be in better shape. And that's some of the stuff we're going to talk about.

and then if things all always have to go up through Mike or Chris or Susan before you do stuff and you have all this stuff flowing up through them you know you've got another problem like this isn't how to run a company and I think we're

looking in the mirror now and these and these operational things these things that were kind of hidden in the old way of doing are very clear and so you know we're we're we're seeing that that command and control or these kinds of things that ego stuff it's it's not going to work in this new world um and let's talk about some of

the other things that happen so when we have unclear priorities everything's a priority this is a big coaching thing for me um and so everything is urgent so nothing's urgent right um and when we have messy communication about what's important and what's not important but we're going to do

that i think but i want you to do this people are feeling tense i mean who's experienced a reorg in the past year of some sort and layoffs and stuff like that right so you're already like where is the end of this runway why am I making all this

change what does this mean for me I worked in I've been laid off four times by the way and I've worked 21 years in corporate before I took off and started doing whatever I'm doing now but the you know the key is is that already has like a bunch of feelings we're carrying that's got a lot of risk we're carrying it's confusing times we need to see what is going on so that we can move forward

and feel good about it and then just like all this change is exhausting because we don't really know what's going on and the changes it the changes about noise and not really having a signal about where to go so what must me do to strengthen as the speed is you know kind of amplifying itself and we

What Better Organizations Are Doing

need to move forward so here's what some of the better organizations are doing and they're not exemplifying all of it so don't feel insecure about it but um

1) Reduce parallel priorities to find the signal

some people are starting to understand they need to reduce all these parallel priorities I mean how many people say we got to do this and then we've got to fix this thing and then we've got to do this thing and do you really know tomorrow morning when you go into work these are the top three highest value highest priority items I need to do who has that list for themselves in the morning yes two three great great I mean that's intense right really seeing the signal

No, those three things are going to matter in 12 months. And other stuff won't. And we don't really know what those things are. So reducing those, really feeling those signals of what the most important is, and getting that overload out.

2) Design decision points and accountability

The second is, where do we need to make decisions as humans inside these cycles? And I would say, if you're a developer and a person that's making an AI tool, really invest in your conversations with your customers about what is going on on their side. and like how are they deciding what is their structure and what's their operating structure this is critical to your ability to scale your product

inside their company hold them accountable see what their investment is and figuring that out showing you what that model is otherwise you're only gonna get to the pond you're never gonna get to the lake or the ocean right and so we really got to understand how things work at places and and they need to reflect on that if they can't answer that be careful because that's another Another place where your product's gonna go in

and you're gonna have a stat like I'm in 50 companies and three say it's great, right? So you should hold them accountable about what they're gonna do about that.

3) Build transparency and trust through learning

And then the next thing is increasing transparency and trust. And what this is really gonna be about these days is about saying, here's what we know, because I don't care what you guys say, nobody's an expert in this room of anything anymore, period, and never will be.

Here's what we know. here's what we're gonna do here's what we're not gonna do and here's why we're doing this stuff now here's why we're doing it having that transparency and taking your team on that journey whatever your job is is awesome and that is actually the way we need to start talking we need to be like vulnerable

about what we don't know you should say here's what I'm learning I went to this event last night I learned about this stuff this is what I did what are you guys learning what podcast are you listening to this is what leadership is is going to look like.

Winning leadership is going to be about intensive talking about learning and understanding what we are and what we aren't to and a bunch of transparency, or we're not going to be on the same ship. OK.

Practical Reflection Prompts to Use Tomorrow

So I'm going to break down each one of these a little bit. And these are just food for thought. You can take pictures if you don't care. You don't have to take pictures.

But I wanted to leave you with these to just be reflection points as an executive coach now. know but I want you to think about like what am I doing that feels really urgent but isn't really that important and it's probably like that report you've got to

get to suit tomorrow or you know these emails you have to send or some of those things like what are what signals are overwhelming you because you're getting too much information I'll just skip over the other one rather than informing you on something that's really gonna make you take action or think about something

like organize yourself here what's going on I talked to a lot of people that are so busy and and busy is a terrible word they're in demand fair but we really need to think about like how am I spending my time and and one of the

tricks I play is is like if I looked at all the stuff that I have on my list tomorrow and all my meetings and stuff what's really gonna matter in 12 months and it's not a lot of it by the way and so we need to be thinking about that

think about this when you're younger, you're gonna make great lagers, I love it. But you know, and what work continues just because it always has, there's a lot of like used to be stuff instead of gonna be stuff.

Like raise your hand and be like why are we doing this? This is a great time to say why are we doing this?

Because if we didn't do this, we could do the new stuff and like figure out these AI workflow stuff and like use base 44 for things. Like this is the world we're in now.

So just a quick question, these are, because we're in an an AI thing I thought I'd give you some prompts but have fun with these like go throw your stuff in there if you've got a big relationship at work with your co -pilot or whatever you're using that

your enterprise has approved go in there and talk about I'm sure you've got a project folder this is all the stuff I'm working what what matters what are the main things I'm doing and and are these the most important things that I'm spending my time on and how would I stack rank these and

like what trade -offs you're like what are the things that are increasing you know my trust in in my team, and what do I need to change around? And then lastly, am I really doing the stuff

that's important, and doing a real analysis on how you're spending your time and what you're doing? I'm gonna leave this open now.

Audience Discussion: Shadow AI, Skills Gaps, and Responsible AI

Any comments or questions? Is this resonating with anybody? I'm seeing some nods in the audience. Yes, oh yeah, totally. Yeah, and I am being recorded right now.

So I will say that even across nonprofit, there's a struggle to look at what they can leverage in AI, what they feel comfortable. They know they have to go in deep if they're going to do that. They know they have responsibility because they're a nonprofit.

I would say in banking, there's places that I executive coach, and they will not allow anybody to use any AI at work. So what do you end up having? shadow AI.

So you go home and there's a Ratatouille TikTok that I, where I get all my news and information by the way, that is the most amazing scene of Ratatouille going around the pot and throwing all the ingredients in and it says, here's me throwing all my company confidential information in here so I can write one email. It's totally what's happening, right?

There's a lot of banana stuff going on and I'm also seeing like just a heaviness even in healthcare pharmaceuticals of how much people are asked to do they almost need the intelligence stuff and then and then and then how do I like prioritize this so

they end up kind of down like a chat hole and they're not even that necessarily good at the prompting and so I'm teaching a lot of that basic stuff like across teams and even inside when you even a technical company I go in and you have to bring like Fran from HR in there and and say here's how you could

you know I'll put an example in there I love breaking news because I learned that from somebody but I go in and say hey like here's a situation a couple of your companies aren't sustainable anymore and that's your promise you know how would you turn this into an opportunity for your board what would would you all, as the C -suite team, marketing and everybody, present to them?

And I first ask them, what are your ideas? Two ideas, right? Total. Like they keep telling me the same idea like five times that are really the same idea,

and then another idea. And so I go and brainstorm tents. They're like, wow.

So they get the tent, and then we go through there, and I say, breaking news. This is what's happened in the board, and here's some surveys from customers. we've gotten back and I want you to instill this and they go through and they learn how to drop

files in there and do all that stuff this is like amazing right so even when they're in there using the tooling I don't think they know what to do I don't think they know how to prioritize doing that stuff and then you have like cool tiny teams where you might have somebody that's like let's get on

this and like start doing workflows but it's so random it's like somebody in a marketing team or somebody's trying to figure out something it's like probably even who you're married to or Or, you know, have you always been into the new, right? And then at the top, there's pressure, as there should be, to figure out what's going on.

Even at tech companies, I have them raise my hand and go, hey, so what school do we go to or certification do we get, to your point, for regulating and, you know, worrying about, like, AI, like, going bad on us? I'm like you need to go educate yourself there's no certification for that like I actually know one certification that somebody gave me but I'm like no this is like your job and I love when like total CTOs ask me that I'm like this is your job to go figure this out I mean it is a hot mess out

there right now and and so I came here I'm like I have a non -technical talk but I'm just telling you like nobody's going anywhere until this part of the house is cleaned up right and so this is is what I'm trying to preach, which is let's start figuring out our position on responsible AI.

Let's start figuring out where these decision points need to sit. Let's start prioritizing that against everything else we have going on in the air. I would say that that should be the top priority of every company right now.

And also, when you think you're not in AI, even on your graph, but it's in every app on your phone. So you are in it.

So if you work somewhere and you use Workday, that is filled with AI. If you're a customer of somebody

and they're using Salesforce, you're filled with AI. So you're already like a sucker for it, whether you think you're on it and using it or not.

Decisions are being made for you. So it's a lot. I don't know if that totally answers,

but I see like kind of this confusion, this like lack of prioritization, this like too much is coming at me. I can't even like do anything with this. There's overwhelm.

And then there's a panic that if we go a little too far, we could lose our jobs. So there's a other whole like change modality thing where people have paralysis, compliance. I'm dead inside, but I'll do whatever you tell me to do with AI.

Resistance and wanting to do things kind of the old way and not so much like catalysts and kind of adapters. And it's a mix. But it's shocking how much resistance is actually still out there. they're out there the resistance is out there because of this it's totally fair

nobody wants to be the person that gets fired for screwing up a surgery I mean I get it and I work with health care people so I understand it so I've given you this to think about and then and then just to let you know like people

Too much work means no room for innovation

that are pursuing too many things which is what everybody's doing everybody's doing too much by the way you don't have room for innovation and thinking no No wonder all of your ideas happen on vacation Saturdays.

It's because you're panicked, like drinking out of a fire hose during the week. And so this is why people always have thought leadership offsites and stuff. It's to get away so they can actually think about something, right?

And people that deliberately prune initiatives and cut them off are more likely to produce breakthroughs. So these are real stats about stuff that's going on. It's all recent stuff.

And then we spent, everybody is ADD now or ADHD now, so that's good, but it doesn't mean we're good at it, right? And we are attention switching and it's killing all this time.

And when I have a lot to do in a day, I mean, it's really like, what was that method I saw on TikTok to make this thing? I was making this thing,

I decided to take a nap on the train, but I was making this thing this morning. like it was bad and then last night and like it was like hard to focus and i kept getting snacks i mean we're exhausted like there's just so much going on um so so thinking about the human

Decision Ownership and Risk: Who’s Responsible When Things Go Wrong?

judgment stuff is that decision stuff there are places where who owns this and who owns this decision is still very unclear even if you're in like a small shop making something and who you you know, who's going to be held accountable?

And I, you know, I don't care what format you use, but like that old racy thing, which is like really annoying and like super creepy and it reminds me of like consulting firms and stuff, is kind of important right now.

Like who is actually responsible? Who's accountable when stuff goes down? I almost feel like from a risk perspective,

if you're going to make any of this stuff, you got to know who's in charge. Even on a tennis team app, like I'm definitely in charge.

And I would be accountable if I gave that to like my tennis team captain and said, try this for a little while. We need to figure that out.

We need to talk about what happens when things escalate. Real mitigation plans need to come forward.

Escalation paths and failure planning (FMEA)

It's almost like this new stuff is making us think about like Six Sigma and like FMEA. Does anybody know what that is? An FMEA, failure modes and effects analysis.

It is the most boring spreadsheet, but I call it a drama queen spreadsheet because it's about like when everything goes wrong, what are you going to to do?

And you break it down thing by thing, what could go wrong? I think we should bring back the drama queen spreadsheet. I mean, it is like time.

So we need to think about that stuff. And I want you to think in your 30 second reflection, what are these decisions that are always like in loops

or like stalls and like Bob has to go talk to Peggy about the thing. And you know, what is that? So, so, and then things escalate unnecessarily. You know how that happens.

So, so think about where this is happening these are things to kind of attack and have open discussions about at your company is this resonating any questions okay all

right and then here are some questions like you know when decisions are looping or escalating you know really what are the assumptions I'm making about what's going on you know think about going in through an accountability lens if this decision goes wrong you know who's responsible and who should rectify that

Sometimes this thing, it's good at guessing at like what people's roles are, if they know what your company's like and what's going on. And then there's like kind of a consequences scan here about, you know, what are unintended consequences?

Here's the best thing I ever heard, and it was at a an event that I went to in DC that was for board people to really start pushing companies on like what their stance is on things and start thinking about things. And this guy who's now in charge of like AI stuff for North Carolina said,

1every single thing we do, I have to ask myself for whom could I fail? And I, and I really think that's an oath, like, that you all should take as you're developing things. Who could you really fail if you do things?

And that gets to bias, it gets to all the things. Think about that as you're developing things. And if there is a failure, who's in charge, okay? okay?

And disintegration and ownership, which is very rare. 20 % of organizations say they have it. And, you know, it really helps people when they have the frameworks built out to understand that when it's high stakes.

Communicating Clearly: Transparency as a Change Strategy

And then last, increasing transparency and trust. I kind of went on this already.

What are we doing? Why are we doing it? And why now?

And why is this like important than other stuff that I might be taking away from you? These are important conversations to happen.

and I want you to think about your life like where would transparency really help alignment for everybody if they really knew what was going on because we always have this like spooky stuff we don't know what people are choosing and why they're choosing we're like doing it until we get told that we're going to be told to go to another direction so like think about that stuff

push up for that manage up on that stuff and think about it and then there's just some things like you know what context do people need when I communicate stuff out that's like a checklist list like what should I always make clear so people don't feel nervous about something

um really talking about that uncertainty and communication and then last like if I were receiving this message way to second guess yourself right what might I question or fear that's a great way to think about your writing and what you're saying especially for technical people that they have been more stem than English but you know it is a mixed bag usually

And then last slide, you know, people that are making trade -offs and doing these kinds of things and being transparent about it, and there's two examples here about newsletters and stuff, these companies are doing great. People feel a little bit less nervous.

In this time, we need more transparency because people are nervous about AI stuff. And then, this is my last slide, don't worry, I'm done. I went fast, right? Okay.

Conclusion: Speed Exposes Ambiguity

Okay, AI is not creating the chaos, right? But it's revealing what our system is doing. And our system is based on systems from like the 50s, right? When we were first making corporations and things like that.

The speed is exposing where we have ambiguous outcomes and solutions and things, and what we need to do is have this clear, this is the path right here.

and if I was making fun little AI tools, this is the questions I'd be asking people when we go to work with them, too.

Like, what's your decision ownership? Who am I going to go to?

Like, how do you guys know what the most important, what are the most important things you guys are trying to get done this year?

How do you guys communicate out what you're doing? You know, who's going to work with me and how much time do you have, right?

And learning that adapts of speed and, you know, the learning thing is like a whole other talk I'll do sometime, time, but this is what we need to think about.

Resources and Closing Invitation

Here's my slide.

I just want to take two things.

I do a weekly AI forum at Chief, and we basically talk about the news, and then somebody comes and demos something cool, like this base 44 thing that you did, and then we record them all, and we put it up here, and then we also pull up other tools and articles that are neat. So this is like a cool hub to always look.

I think we have like an NA at a like demo, and people are playing with quad bots, and making bots of themselves for LinkedIn and stuff so we're doing like kind of fun stuff but if anybody wants to come present we would love for you to

do that it's every week on Thursdays at lunch and then please connect to me and I also go into corporations and like talk to the non -techs and the techs about how we can come along in this process so feel free to reach out if you need any training thank you