The Coming Wave of Ambient AI

Are you ready for AI that’s intuitive enough that you don’t even have to tell it what to do?

How about AI that can anticipate the right next moves?

Today, we’ll share how ambient AI is going to change your business.

Hi, I’m Courtney Baker, and this is AI Knowhow from Knownwell, helping you reimagine your business in the AI era.

As always, I’m joined by Knownwell CEO, David DeWolf, Chief Product and Technology Officer, Mohan Rao, and Chief Strategy Officer, Pete Buer.

We also have a discussion with Dan Chuparkoff about why AI is like a pinata and how it’s going to change work.

But first, here’s my recent conversation with David DeWolf and Mohan Rao about the coming wave of ambient AI products.

I want to read you something that David wrote at the end of last year in his predictions of what to expect in 2024.

David, you said, the UX of AI will become more sophisticated and much more seamlessly integrated into our daily workflows.

Rather than users going to AI to ask for insights, for example, AI will begin delivering insights without being prompted.

So without dissecting whether or not your prediction has actually come to fruition yet, let’s use that as the leaping off point.

What do we mean about ambient AI and what do people need to understand about where we all believe we are moving towards?

I think this is actually an extension of a trend that we’ve seen for a very long time, which is technology as it progresses blends into our lives more and more.

I love the example.

It’s so simple.

We started computers were mainframes, and then we got desktops, and then we got laptops, and then we got mobile phones, and then the mobile phones became watches.

It’s just getting smaller and smaller.

It’s blending in more and more.

We see the digital screens everywhere that are part of our environment.

It blends into the wall.

It’s where the picture is, but they’re giving us information at the same time.

That is the trend.

If you take what AI is able to do now, it is able to not just present something that was predetermined.

The predictions, the forward thinking aspects can really start to learn me personally, can learn you personally, and can begin to take action on its own, to insert this technology into our day-to-day lives.

I really think that even what we consider modern technology platforms, SaaS platforms, like in the Cloud, I go and I use the software somewhere, I think that’s a stale paradigm.

I think we are moving to a world where it’s not just the physical device, but even the virtual software is going to be in real time.

You and I are having a conversation and all of a sudden, it speaks up and it corrects me because what I said was wrong, right?

Or we’re able to invite it into the conversation and ask it a question because we don’t know, right?

And there’s going to be much more fluid interaction with technology versus me having to disrupt my life to go to a screen and play with something.

To me, that’s ambient user experience.

Is it blending into my life and being part of my normal, typical interactions and me not having to interrupt my life to go interact with it?

A good, good early example of this is my Gorman watch, right?

I wear it.

We all wear my watch or one of those.

It’s there.

You think of it as a watch.

But when I’m stressed, it lets me know.

When my energy level is dipping, it lets me know, right?

So what it is doing is sitting there in the background, it’s unobtrusive, operates in the background without requiring any engagement from me.

But then comes to me and beeps and says, hey, you appear stressed.

Just maybe you should go for a walk or go get a drink, something, right?

So that’s a good example to think, and then you will have to then extrapolate it into how, what would this mean in a business context.

Mohan, you’ve talked a lot about knowledge management in the business context.

Do you see this ambient UX and knowledge management within an organization kind of intersecting more quickly than maybe some other things?

Yeah, absolutely.

I think when you have AI-based knowledge management, as opposed to the watch example that I said, where it lets me know, it beeps that I need to move or I need to drink some water or whatever.

In a business context, Courtney, what this can do is perform certain tasks, right?

So it’s able to sense a situation and be able to perform a task, and furthermore, from there, make a decision and just learn from the context that it’s operating in.

Knowledge management is critical to that in a business environment from learning the context and having the knowledge base to make decisions.

Here’s what I want, Mohan.

You tell me if I can have this.

I want to go to a meeting, a Zoom meeting or Google Meet, whatever it’s called on Google.

You want to go somewhere.

I’m going to a virtual meeting.

I want to get out of that meeting, and I want something to just be like, hey, here’s your three action items.

Do you want me to go execute these for you?

Like, it looks like you need a 15 minute with, you know, a so and so, da, da, da.

Like, just have all the things, and be like, good, go.

Oh, I think it’s going to go even further.

Am I going to get that?

I think you’re going to be in the moment, and it’s going to schedule that meeting as it comes up in the meeting.

And I think it can actually predict what meetings you need, right?

I think it will go even further than having to ask you.

What if I don’t want it to do that, though?

Well, then you’re going to have to talk to it, and you’re going to have to tell it, just like you would any other employee.

Okay, okay, I like this and kind of don’t like this, but I’m here for it.

Okay, so you think that’s going to be possible.

What else do you think is going to be possible that I haven’t dreamed up yet in the business context for Ambien AI?

Well, I’ll give you a real live example that we’re building at Knownwell, right?

As we are building our commercial intelligence over time, we will be able to take action on that intelligence in order to strengthen customer relationships with and without human interaction, right?

And there is a corpus of knowledge that we need to build up to be able to get there, but we are on the path to being able to automate, right?

Take the example, you use the example of sitting in a meeting and talking about, we need to get together afterwards and talk about this and it automatically does it.

Well, think about connecting dots, human occasion that’s already happening across the organization and coming to a judgment and a conclusion that, you know what, something’s going wrong in this commercial relationship, and so I’m going to get two parties together proactively to talk about this before they even know they have to talk about it, right?

Those types of things in a business context, I think are going to be normal in orchestrating the business to be more efficient together.

Okay, I love that.

In our last episode, we talked about native AI technology versus AI infused technology, and we were like, hey, what’s falling over here?

What’s the difference?

What you just described, I would assume, correct me here, that that would only be possible for AI native technologies, or do you think that’s going to be something these add on?

Courtney, maybe I can give a sense of what it will take to build something like this.

I love that.

Just provide those building blocks, because it’s really hard to answer your question of what can this do.

You can almost imagine anything that this can do.

The potential is that vast and it’s hard to build.

Right now, when most people think of AI, Gen.AI, they think of a chat interface.

That is the predominant interface.

It is an useful interface because there is a human in the loop that is gently correcting your AI pal on the other side.

When this matures, from streaming chat, we’ll have an equivalent of a non-streaming chat.

It is not words that are just coming at you, but it’s happening in the background.

What that allows you to build are these background agents.

And these background agents that are running in a non-streaming mode, and they will ask for help when they need help.

That’s gonna be an interesting problem to solve, but knowing when to ask for help, that is the intelligence that they need.

So where we are going in the future is human in the loop to human managing the loop, and from there to human delegating to the loop, right?

So these are the stages that need to happen for us to truly realize the potential of ambient UI.

As they go into more delegation, you can then observe what happened.

You can rewind, you can edit, you can rerun, but it’s this transformation from in the loop to managing the loop to delegating to the loop that needs to happen over time.

And that’s when we’ll have this ambient UX.

This is a whole new world.

I do, it is fun to think about.

Do either of you have a time horizon of when we might expect this, well, from my opinion, glorious moment in time?

When are we getting this?

I think this is an evolution, not a revolution.

I think, you know, we pointed out some examples at the beginning of the Garmin Watch, the Alexa.

Like we’re already seeing early phases of it.

I think you’re going to see that mature and grow over time.

I don’t think it’s going to be a aha moment, but I do think in a year or two, we’re going to start to see a lot of this in software in a way we have seen it in hardware.

I feel like there’s lots of consumer driven instances that feel like they’re maturing faster.

When are we going to get it in the side of the world that’s making the money?

When are we going to get it in business?

I don’t think the time is that far away here.

You will start the initial applications of this starting to roll out as early as next year.

It’s not that far away, but it’s going to be the types of tasks that you’re going to trust that you can delegate.

Most of the initiative here has to be on the humans for us to trust and there is enough experimentation that’s gone on, but I can see the leading edge of this happening relatively soon.

Is this going to be like, obviously, I’ve used Alexa a lot, but I know a lot of people that will not.

Is that going to become a problem when this type of technology is infused into our business lives and they’re not able to trust?

Yeah, I mean, absolutely, because legally, you’re still responsible for your actions, whether you do it or an agent does on your behalf.

So we’ll have to find use cases initially, where it’s low risk from a privacy perspective, from a legal perspective to get this into motion and then build up that trust and understand how to build the safeguards around privacy, security, legal aspects before this becomes widespread.

I think there’s also going to be a bit of a paradigm shift.

And you’re starting to see this with content and authenticating that content is real, where instead of a default assumption of trust, there is a default assumption of fake, that then you validate.

I wonder how much that will start to play into these types of things we’re talking about, is that there’s just going to be a new paradigm about how we think about things.

That’s really interesting.

In the interim, until we have that, Mohan, are you and I just going to be able to anytime we mess up, which doesn’t happen because you and I are both perfect.

But if we did mess up, we would just be able to be like, David, it wasn’t us, it was that ambient technology doing things for me.

I don’t know.

I used that excuse already.

I didn’t send that e-mail.

I don’t know who sent that.

These AI agents are really mischievous.

Oh, that’s funny.

Oh my God.

Okay.

David, Mohan, thank you for this education about ambient AI.

Hopefully, as you’re listening, you’re excited, maybe a little bit, and see maybe some applications for this coming on the horizon soon.

Thank you too.

Thank you.

So much fun.

Thanks, Courtney.

Thanks, David.

Y’all, this episode is going live Monday, September 23rd.

And if you’re listening to it on Monday, September 23rd, tomorrow we have some really big Knownwell news to share.

So if this is on Tuesday or afterward, we would love for you to go to knownwell.com/news to find out what the big news is.

We are really excited to share it.

And we also have a really exciting LinkedIn live session coming on September 27th, where you’ll get an inside look at what we’re building at Knownwell.

And join David DeWolf and I for a Q&A about our big news.

It’s going to be a lot of fun.

So again, tomorrow on Tuesday, this week, we’ve got really big news.

We hope that you will follow along and see our big announcement.

And I hope to see you on our LinkedIn live on Friday.

Go to knownwell.com to stay up to date on both.

Pete Buer joins us this week for a new segment.

We’re calling, How Much is That Software in the Window?

I’m personally very curious to see how it shakes out.

But Canva says its AI features are worth 300% price increase, according to The Verge.

Pete, what do you take away from this story?

So I can’t help but think there is something going on here that I don’t fully understand, but at least in terms of what the article is able to provide.

It’s a bit of a head scratcher to me.

On the one hand, the target market for Canva is, in fact, the place where AI-powered work enablement tools have a home and drive real returns.

On the other hand, an awful lot of incumbent platform companies have been launching their AI bolt-on services and products for free as part of the foundational platform pricing.

And so this approach stands out, at least relative to the pattern across the rest of corporate society.

On the third hand, if there is such a thing, from the article, it sounds like the communication around pricing has been a bit ham handed.

Like, can you imagine getting the email that says, without any explanation or any human interaction, your price just went up by 300%.

And by the way, all of this for a company that’s established a reputation for user-friendly, buyer-friendly pricing.

So I guess the new AI-powered offerings must be really spectacular.

And maybe with that, I’ll volley back to you, Courtney.

You and the team are Canva users.

What’s your take?

We are, which is, it’s really interesting because, frankly, I’ve been a fan of Canva, especially with a startup like ours, it’s allowed us to do some things in-house that we wouldn’t normally and we’ve gotten to use some of the AI tools firsthand.

Now, they may be planning to roll out some newer ones, but what made it so appealing was the affordability of it.

You know, I’ve always been an Adobe user and fan and definitely recommend it for designers still, but it was the price point and accessibility that Canva provided for the layman user on the team.

To just be able to turn out something for a slide or something that they were needing pretty quickly.

But, you know, a 300 percent price increase might make me second guess, like, hey, do we just go back to, you know, a Photoshop product on the cloud instead of Canva?

So I think it’s a really interesting.

I’m wondering if the price increase will actually take them out of the target market that they’ve been able to own almost completely and open it up for some smaller, maybe some newer companies to come into that space that they’ve created.

But bold move.

And you love the notion of attaching real value to the power that AI brings to making the business more efficient, more effective.

But to your point, are they boxing themselves out of what is otherwise a market that they could own?

Yeah.

And we can obviously say firsthand, AI technology is not inexpensive by any means.

But I think for some reason, we’ve positioned it and so many platforms are just tacking it in.

When you go to Google, now when you search, you’ve got an AI tool right there at the top working for free for you.

And so I think it may be a rude awakening for us when the cost of that technology starts to get passed down to us, the consumer.

That’s going to be an interesting storyline to follow as it matures.

Totally.

Well, Pete, we will help all of our listeners stay in the know when it comes to these price changes and how AI is changing products that we’re using in our businesses.

Thank you as always.

Thank you, Courtney.

Dan Chuparkoff is an AI and Innovation Keynote Speaker who has worked as a technology leader at Google, McKinsey, Alassian and more.

He sat down with Pete Buer recently to talk about why AI is like a piñata and how it will change the way we work.

Dan, welcome to the show, it’s so great to have you.

Thanks Pete, it’s great to be here.

I’m looking forward to our conversation.

Same here.

Your background is awfully impressive, companies like Google and McKinsey.

What are you doing currently and how does AI factor in?

Yeah.

I quit my job at Google last September because little by little, I started talking more and more at conferences about tackling AI, the new AI transformation.

I realized I’m good at that.

I’m helping people and they like my help.

While at Google, I could only help the team that reported to me.

I decided to quit my job and go out into the world and help people figure out what AI is.

Is it a bunch of hype?

Should we wait for the dust to settle or the bubble to burst?

Or should we dive in next Tuesday and figure out what to do with it?

That’s what I do full-time now and I love it.

It’s great.

It’s the best job in the world.

I hope the answer isn’t and it depends, but you just laid out a continuum that runs from hype to get started next week.

Where are you generally on that continuum?

I think the first thing I generally help people do is to break the word AI into various components.

Because AI is an umbrella term that means a lot of things to a lot of different people.

On the one end of the spectrum, you have like heavy, deep machine learning neural net scientists.

And on the other end, you have like my nine-year-old nephew, Connor, like making a picture in mid-journey of a zombie riding up a unicorn.

Those aren’t the same thing.

And so I help people break things apart, break AI apart, understand what’s sort of theoretical and what’s practical now.

And the most practical thing right now is AI note-taking.

If you’re having meetings with a bunch of people and talking about a bunch of really important stuff, and you’re not using an AI note-taker of some kind yet, you’re probably missing out on some amazing immediate-term value.

What’s the role that AI is going to play in reinventing the way teams work?

Yeah, I think it’s, I would tackle that in the opposite direction.

First, I’ll say it says reinvent on my shirt.

In a single day, I have like 18 of these things.

And I do that because I think that people need a constant reminder that the world is constantly changing.

And so the way we do things needs to constantly change as well.

And so instead of saying AI is going to reinvent the team, what I actually believe is if you want AI to be a part of your future, you’re going to have to reinvent the way you learn things, try experiments, share information, make decisions and apply new technology in your team.

Those are the five things that I think are the main things that power change inside an organization.

And so I have a new book coming out.

It’ll come out in Q1 of next year, and it’s called The Innovation Machine.

And it talks about reinventing those five core strengths of an organization.

When you figure out how to constantly evolve those things, then tackling remote work isn’t as hard, or tackling AI isn’t as hard, or tackling robotics, that’ll be the transition that comes five, six years from now.

And all those things require reinventing the way you approach your day-to-day conversations at work.

As onlookers to this industry, we spend a lot of time doing research, trying to understand the impact of AI and which industries will be affected the most and how.

And a lot of the data that you get is finely tuned assessments of the degree to which roles, jobs, teams, work will change.

So what percent augmented?

What percent automated?

I know that you’ve got a framework, the hierarchy of human experience, that I believe speaks to at the end of the day, the jobs that are best done by humans.

Can you tell us more?

Yeah, yeah, of course.

I think that there are some things that AI does really well, and some things that AI struggles with.

And it struggles with those things because those things are different every time.

There aren’t enough examples in the training data, right?

And so the more unique the thing you’re doing as a task at work is, the harder that will ever be for AI to do, even as AI gets more and more and more sophisticated, a brand new problem that the world has never seen before can’t be solved by a thing that requires training data to make it work.

So I have what I call the hierarchy of human experience, and what that refers to is the series of tasks that you do from, you know, in your day-to-day life.

And at the bottom of that layer, at the bottom of that pyramid of work is process, just processing tasks, doing things over and over again, following instructions.

Then you have communication, talking about doing those tasks.

That’s the second layer.

Third layer is investigating, looking for issues.

Those are the first three kinds of work and most of us spend a lot of time doing those things.

AI also can do those things, but there are three more levels of work.

The fourth layer of work is problem solving, because once you investigate issues, then you start finding some.

Problem solving is a really uniquely human thing, because we’re willing to take risks and try things that were never tried before.

You have to do that when you’re solving a brand new problem.

And so as soon as you get to that problem solving layer, I think you’re in a uniquely human thing.

So once you’re doing some problem solving at work, I think you’re safe for a long time.

But it’s important that you give away some of the other stuff so you can start solving harder problems, start solving more problems.

There are two more levels of work, discovery and imagining.

And those are also important, but I don’t want to belabor the point.

That problem solving threshold is the really important part.

We’ve seen some data recently from Microsoft saying the majority of employees are using generative AI at home.

They’re bringing AI to work, I think was the language that they’re using.

But that half of them, 52-54%, are uncomfortable admitting to the fact that they’re using it in some of their most important work.

How big of a problem is this and what should leadership do?

I think some amount of that goes away at some point.

Like, really, imagine 15 years ago, and you’re writing a word document or an email to a colleague, and you don’t know that spell check exists, right?

And so you’re trying to spell all your I before E’s, right?

All those things, right?

You’re there, there, there’s, you’re your, you’re yours.

Those things we used to have to do manually, and then spell checks started nudging us along and helping us out.

Spell check doesn’t feel like cheating anymore.

It’s so built into every single tool that we use that it’s, it’s just like the fabric of the universe now, and AI will be like that too.

But the second thing about it is, AI, fundamentally, because AI draws from the populous, like, it draws from the most common answers to questions, by definition, it’s doing average work, right?

And so, if you were doing below average work, and now you’re doing average work by using ChatGPT, that’s better, and you should do it.

But it’s uncomfortable realizing that, oh, the email that I was writing before kind of sucked, and now it’s better, and that’s in my face.

You should use it.

It will make your stuff better, especially the commodity stuff like sales outreach emails and blog post intros and all those things.

It will make it better.

But also, there should be something in your day that requires problem solving and expert level work.

If you’re doing just average level work or below all the time, your job is kind of a commodity and it will get automated, right?

And so be cognizant of how much you’re leveraging it and what you’re bringing to the table beyond, beyond ChachiPT, because that’s an important question for everyone to face.

The reluctance of users to admit that they’re bringing their tools from home or that their work, you know, has had an input before, you know, they got their pen out.

Concerns me, especially in light of the fact that what I’m hearing from you is that it’s all going to get better.

This is, you know, we’re on a path here.

We just need to get to a place of comfort and maturity with it.

What can leadership teams do to, I don’t know, eliminate the fear and encourage the experimentation and let the wave crash a little more naturally?

Yeah, yeah.

So first, I think organizations and administrators and, you know, tool deployment people should recognize that a lot of this isn’t fear, it’s impatience.

Right?

It’s that, hey, I can use GPT and it will make me better, but you won’t let us use it yet because you’re worried about the, like, implementation or the cost or the risk or the, like, all of those factors are slowing organizations down from rolling it out officially and your people aren’t going to wait for you.

So first, aggressively figure out how to speed that up because your organization has AI already.

So stop trying to pretend that you’re mitigating the risk by not rolling it out because it’s out.

And have a plan for being AI mature by the end of 2025 because your people are bringing it there, whether your procurement people are ready or not.

Thank you for sharing your nuggets of wisdom on the impact that AI is having is going to have on business and frameworks for thinking about it.

I think it’s been hugely helpful and thank you.

Thanks Pete.

Thanks as always for listening and watching.

Don’t forget to give us a five-star review.

It really helps more people find this show.

At the end of every episode, we like to ask one of our AI friends to weigh in on the topic at hand.

Hey, Jim and I, what’s happening?

This episode, we’re talking about ambient AI.

What do you think is the future of ambient AI in business?

Ambient AI has a bright future in business.

Imagine having AI assistants that anticipate your needs and streamline your workflow.

It’s like having a personal assistant on steroids.

And now, you’re in the know.

Thanks as always for listening.

We’ll see you next week with more AI applications, discussions and experts.