InterConnect 2017 Conversations: Steve Ardire talks with Dez Blanchfield

Screen Shot 2017-04-05 at 12.19.35 am copy

Steve Ardire is the gentleman on the far right of this photo ( to my left as it were )

 

The following is a transcript of a “fireside chat” podcast with Steve Ardire recorded at IBM InterConnect 2017 in Las Vegas ( USA ).

Listen to the podcast here => http://j.mp/IBMInterConnect2017SteveArdire

Dez Blanchfield:

Welcome to the IBM InterConnect 2017 podcast series. We’re coming to you from sunny Las Vegas. I’m your host, Dez Blanchfield, and I have the pleasure of being joined by Steve Ardire.

Now, I’m going to quickly introduce you, Steve. You’re a regular tweeter. You’ve been online since 2008. I see from your profile you’ve got a solid following of about forty-one hundred followers.

You describe yourself as a strategic advisor focusing 100% on AI startups. You live in Olympic Peninsula outside Olympic, Washington State here in the USA and there’s a quote that I grabbed from your profile.

It says that you are, in one line, you “connect and illuminate the dots that matter for business strategy, funding, go-to-market, customer and partner engagement.” Welcome to the podcast. How are you enjoying the show?

Steve Ardire:

Thank you, Dez for that very tight, succinct introduction and that pretty much depicts what I do and operate. How am I enjoying the conference? I was at both InterConnect 2016 and World of Watson.

My takeaway from the high-level view is I was very impressed with, the messages were, the key messages in terms of keeping data simple, getting inspired, empowering teams.

We’ll focus on that one a little bit further and in building hybrid cloud, we’re seeing the manifestation.

My overall takeaway is that I’m seeing real instances of IBM delivering on the promise of cognitive computing with nice, tighter synergy. Of course, it needs to get tighter further, but across the cloud story to IoT and so on. That’s impressive.

Dez Blanchfield:      

If you were going to introduce yourself in a speed dating form, how would you describe yourself in 60 seconds?

Steve Ardire:

Okay. You got the thumbnail and I’ll just hit the point lines on that. I’ve been an advisor for 20 plus years. The focus was across different startups. Four years ago, I made the concerted effort, kind of I saw this tsunami coming and I said I’m going to get, well, maybe not a tsunami.

A big wave because you’re a surfer from Australia, right?

Dez Blanchfield:      

A big wave. Everyone surfs in Australia.

Steve Ardire:

Exactly, so I saw the big wave come in and said, “I’m going to get on this AI wave,” and that’s exactly what I did. I really dialed into … My expertise is really working with early stage software startups.

What I look for is really how to, as you describe, how to shape the business strategy, the messaging and get funding, go to market and tie them to minimally one partner ecosystem.

The ones in my current lineup of six AI startups, two of them are IBM Watson plays. We’ll talk about it more in terms of what I met up with in terms of financial services health care, which is really kind of a successful outcome of delivering on what I espouse to do as an AI advisor and a merchant of light.

Dez Blanchfield:      

Well, that’s actually a nice segway into my next question for you, and that was what was your biggest OMG moment from the event so far?

When we were talking before, I hit record, you mentioned that you saw the Watson financial services team here in Las Vegas demonstrate how they’re delivering cognitive services in the key vertical market around financial services in the industry in particular.

What can you tell us about what you saw there and your general takeaways from it?

Steve Ardire:

Exactly. This was a new service evidently. It was just architected in like a couple of months to get ready for InterConnect, which is very impressive.

Dez Blanchfield:      

They admitted they weren’t coders, right?

Steve Ardire:

They were not coders. This was really taking kind of functional architecture, some Python scripting and putting it together, but the most impressive piece is the fusion between the IBM Bluemix and functional microservices to be able to classify, categorise, package together.

It’s really showing how to create the holistic architecture of data aggregation enrichment, standing up application services with microservices.

We saw it there and it was very, very impressive. That was one of my key takeaways. Furthermore, I leveraged it even a bit further. I said-

Dez Blanchfield:      

Of course you did.

Steve Ardire:

Yes. Of course I did.

Dez Blanchfield:      

I’d be surprised if you didn’t.

Steve Ardire:

Exactly. The thing that was really interesting. I said, “You know, in AI, most credible AI players, Watson and others, can address the who, what, where, how and that’s important,” but I said, “You know,” when I was looking at what they’re doing in this IBM cloud financial service, I said, “You know, what’s really the missing ingredient and the holy grail is causal reasoning with context and time”.

When I played that out, this is not a pitch for the startup, one of the startups I represent called Meelo.com, they said, “Wow. I get what you’re saying here,” so we took that conversation to the financial cloud after party and had a very convivial time and there’s going to be a follow-up meeting to discuss this further.

Dez Blanchfield:      

Right. I remember in fact there was a funny little segue here. A young guy was in a purple jacket and we asked him about his purple jacket and he said that he was running around town on a particular meeting.

He gets in the cab and the cabbie’s like, looks at his purple jacket and says, “Respect.” And then hits the metre and was like, “Whoa, dude. Full respect,” and it turns out it was the day that Prince passed away.

Steve Ardire:

Exactly. That was phenomenal.

Dez Blanchfield:

Phenomenal right. He was a snappy dresser. He said something that really stuck with me, and I actually wrote it down when I went back to my suite last night.

He said that they weren’t coders and they developed to APIs and I think for me that was quite a significant takeaway, that non-techy, non-developer people can Grok APIs and grasp the concepts how to plug into microservices as you said on Bluemix and consume the likes of cognitive and Watson as a service.

Do you think that is the trend now? Do you think that we’re seeing people come from a non-technical pedigree who can Grok technology and put it to good use for their own purposes? Is that what we’re seeing out there now as a trend?

Steve Ardire:

I think that is spot on. It’s the trend that … Because first and foremost, data scientists don’t scale, period. By-

Dez Blanchfield:

My stomach might argue with you, but anyway.

Steve Ardire:

But I mean, sure. There’s always the need, but in terms of scalability, this was a excellent example of being able to stand up a functional, impressive, very useful applications that’s taking the infrastructure services, middleware services, in cloud, others.

It’s really coalescing all of that into something to where, hey, you’re lessening the dependencies on taxonomists, ontologists, data scientists, which is a good thing and Joe and Jane Knowledge Worker are going to profit from this as these new services for connector services.

I saw, second to this, I saw something very impressive in health care. The product manager for IBM Healthcare is putting in patient care where they’re using text analysis, leveraging the Alchemy and actually what they have done, which is very impressive, they’re taking specialised coding in medical and embedding that into this version of NLP, so you can take medical records and they are working on, with the big providers there, to be able to merge.

In other words, in a much faster fashion, be able to ingest that data, come out with what the patient records, these are the key causes, the concepts that you should focus on, so the doctor has faster time to insights to be able to address that. I saw indications of that. They’re also maybe at a similar juncture as far as starting to roll that into the marketplace, which I think is terrific.

Dez Blanchfield:

You mentioned something that’s quite hot at the moment and I’m going to let you explain it. NLP. Explain the acronym.

Explain the significance of it because I think folk listening in may have kind of missed that salient point. I think it’s a really critical thing to touch on. Just a quick 30 second on NLP and kind of where you see it being a critical component of-

Steve Ardire:

Right, right.

Dez Blanchfield:      

Watson and Natural Language Processing.

Steve Ardire: Sure. NLP stands for natural language processing and there’s also a corollary called natural language understanding, which is a little more advanced. The whole point of, it’s so as you ingest, as you take text analytics, and that’s essentially what’s built in, one of the components of Alchemy has an NLP component.

The whole point of it is like in the IBM conversation manager, the Watson conversation manager, they have, we even heard it here at the show, 95% accuracy as far as going speech to text, which is impressive.

There’s text to speech, there’s speech to text, but the holy grail of all of this is to understand more of the nuances of behaviour and intent. This is getting us to what all the big players, IBM and others, are doing is conversational understanding.

That’s the holy grail where we’re getting closer to that, but we’re seeing really positive breakthroughs along the way.

Dez Blanchfield:      

Ginni Rometty impressed me more than usual at World of Watson last year when she got on stage in front of more than 20,000 people and for nearly an hour-and-a-half, spoke about one of the biggest companies on the planet and some of the most complex technologies on the planet, but she didn’t geek out once.

The five key things I took away with it were all humanities: medical, science, whatever, manufacturing, music. I think this is where language processing becomes a big thing.

Yesterday, in her CEO and Chairman’s keynote, as you just outlined, she quoted a couple of stats that I took away as well and then she said, “Look, Humans make roughly 5% of error on anything they do, particularly in language processing text, right?” Apparently she said that Watson’s at 5.5% accuracy of error rate.

That’s like less, well let’s say half a percent away from being supposedly human, although I think we often conjure up images of what an AI looks like and it’s often a robotic format, but that’s getting pretty damned close to being so clever that you almost can’t determine the difference of what’s on the other end. It’s like the Turing-test.

Steve Ardire:

Exactly.

Dez Blanchfield:      

All the right stuff, which is interesting. I guess it’s how we put it to use, then, is the big question, isn’t it?

Steve Ardire:

It’s really is. Look at the ramifications. This is why it’s kind of a continuum. You’re observing. If you have the vision recognition where you’re understanding facial recognition and emotion detection and then you have in a ambient form in the IoT where you can, so you observe, you listen, you have the sensors out there.

You’re starting to see industries like in automation for oil and gas, which is my first career, the whole point of this is really allowing people now to focus more on the critical tasks. Like in the enterprise today, systems of intelligence are replacing systems of record finally.

That’s impressive because one of my favourite quotes that I always use in presentations is alluding to Peter Drucker, the management guru, “Effectiveness should be a human pursuit and efficiency should be delegated to machines.”

As we’re getting the adoption of AI across the board, here’s three key points, and I’ll close on this for this segment, is machines can attend to vastly more information and more complex processes than humans.

Yes, many jobs will be automated, but this will drive the human capital to hirer order, non-routine cognitive work, but the nexus and we heard it here at the show, it’s really … I don’t even use machine intelligence anymore. My favourite is augmented intelligence because really, it’s a matter of where machines and humans work together and the outcome of that is things like faster time to insights, et cetera, et cetera.

Dez Blanchfield:      

I think if we look at it from a traditional enterprise value proposition, what I constantly get asked for and I’m sure you do as well is, “How do I just reduce the time to value?”

Steve Ardire:

Correct.

Dez Blanchfield:

“How do I essentially make money out of this?” in many cases, but not always. Sometimes it’s, “How do I make my staff happier? How do I make my customers happier?” The whole celebrity experience thing is just completely disruptive. What MarTech, AdTech sales, et cetera. Sports ..

Steve Ardire:

It is. The other thing corollary to that, Dez, is really empowerment. People work at a higher … We’ve talked about this..

Dez Blanchfield:

Yes, indeed.

Steve Ardire:

During the.. People feel more empowered and where they, you’re depending on them. Yeah, I have confidence in you. You can take the ball and run with it and I’m confident you can come up with a positive outcome with more minimal guidance, which is terrific. Rather than just having overbearing managers, we’ve all been through that before.

Dez Blanchfield:      

My gosh, yes.

Steve Ardire:

It says, “Okay, if you’re really not adding value as a manager, what the hell are you doing there?”

Dez Blanchfield:      

Yeah, definitely. I think I’ve seen a real shift where managers who are successful can provide the tools to let their staff gain that self-empowerment, that self-duration particularly when it’s the tools from folk like IBM where you’ve got data-driven decisions.

Steve Ardire:

Right.

Dez Blanchfield:

Staff walking around with large format tablets with dashboards and making real-time decisions that are … You know what I mean?

Steve Ardire:

Exactly. This is wonderful because remember, for years, CEOs had said, “I felt it in my gut.” That’ll be fine and well, but for heaven’s sakes, obviously you’re not a data-driven organisation.

Dez Blanchfield:

No, no. Too often, even if companies thought they were data-driven, they were looking at PowerPoint presentations or spreadsheets that were from last month.

Steve Ardire:

Correct.

Dez Blanchfield:

How can I work to the future if I’m only getting historical data that’s a month old?

Steve Ardire:

Exactly. I’m glad you brought up that point because really I’ve used this before. I mean, right now, the future of business is flow.

It’s getting faster. It’s getting more real time and the complexity is getting enormous, so really you’re getting higher and higher degree of heterogeneity in your data sets from webs, from streams in your enterprise systems and it’s a matter of if you can apply machine intelligence to be able to glean and process text analytics, understanding along the way.

This relieves the burden of being able to classify, categorise, to get the enrichment. It’s given you faster actionable data so you can make those decisions faster and smarter.

Dez Blanchfield:      

I do like those. I’ve just written them down as you saw, but just so the folk who are listening, I like the idea of you describing it as new flows. Speed is a given, but it’s cool to call it out and complexity, I think that those three things wrap up into the kind of the new working model.

To move onto another topic now. The three key things that I have sort of taken away from this event that’s really being punched forward are enterprise strong, data first and cognitive at the core.

I’m really keen, if you don’t mind, if we could kind of like do some rapid-fire, 30-second responses to each of those. Is that okay?

Steve Ardire:

Sure.

Dez Blanchfield:

Let me run them past you one at a time and just get your 30-second rapid-fire response. In 30 seconds, what’s your takeaway to folk listening in to the idea of enterprise strong? What does that mean to you based on what you’ve seen so far in the last couple of days?

Steve Ardire:

Enterprise strong, and I touch upon it earlier, is the systems of the intelligence replacing systems of record. Systems of record is only as good as the quality of the human input, so you really … They’re relying on human curated rules, rules that are by definition static.

They become obsolete, so the whole point of machine, of re-engineering is that you’re creating new workflows where it has the potential to augment on thinking to have better understanding, improve operations, especially the more complex operations, so the whole point of it is you’re using this software so companies continue to learn from the data as conditions change.

Dez Blanchfield:

Awesome. Now to second one. Data first.

Steve Ardire:

Data first. This ties to some of the other thoughts regarding being data-driven strategy for the reasons we’ve already cited, is that you’re lessening dependencies on human curators.

You’re using machines to do that. The whole point of a lot of this is, this is one of my quotes that I prepared for our group here, the social analytics, is, “Machine intelligence is a prediction technology.

As it improves, human prediction skills will decrease. However, the value of critical thinking and human judgment will increase.”

Dez Blanchfield:

I love it. Third one and for me, this is I think going to be a longer tail version, topic, rather, in that it’s going to take a while for companies to really grasp the value proposition that it offers and consumers probably won’t actually see it because it’ll be under the hood, but in 30 seconds rapid-fire, cognitive to the core. What does that mean to you?

Steve Ardire:

Cognitive to the core is really kind of this continuum I was talking about is that as far as having integration from data first, the enrichment in the cloud, the APIs driven with micro-services where you’re really improving the business processes where you’re getting the transformation of AI to augmented intelligence, frankly.

Dez Blanchfield:

Right, right. I wrote down a couple points that link to that where you sort of mention the idea of machine to human and human to machine and I think that really links in nicely what you just mentioned there.

I’ve actually got a screenshot that I took from the keynote we just attended a while back this morning around the data first method. I’d like to just throw a couple questions at you.

They describe it as being a methodology or a framework that flows from an initial briefing and vision of what you’re setting out to do, a discovery workshop, a design and validation process, implementation, and I imagine there’s sort of an iterative process there, and then run and maintain.

Is that what you’re kind of seeing? Have you seen that come about as far as that methodology goes in other ways though as far as the agile approach to data? Do you think that’s a significant new innovation or-

Steve Ardire:

I think it’s an enhancement. More than enhancement, a healthy enhancement because really the whole point of is that the tools are ever-expanding in terms of data aggregation and enrichment. Then we’re getting much more sophisticated instantiations of deep learning correlation processes.

That is identifying what’s happening. Then you can process that with the methodology we talked about, but I really think that this last part with causal AI is very, very important. Causal AI, which determines why things are happening, has the greatest potential to change how industries operate, businesses compete.

I’m very keen on that, which is why I look for opportunities to interleave like a Meelo.com into financial services, into media. To be able to AI with hyper personalization, you’re really getting down to understanding … I mean, this is a marketer’s wet dream-

Dez Blanchfield:

Yeah, absolutely.

Steve Ardire:

To understand, “Okay. If I know this person, I could not just sell them stuff, but understand what are their key triggers? What motivates them? How I can …” That’s really building this customer enhancement to a whole nother level, which is really wonderful.

The whole point of this is the other theme I’m really keen on, just as a throw-in, is I really think AI has become the new UIX. With this hyper personalization, the IoT stuff with the sensors and so forth kind of disappears in the background, but it’s there to help enhance the customer, the user experience.

Dez Blanchfield:

The amount of data we’re producing now is so beyond, so far beyond not just human comprehension, but human capacity to absorb and deal with. We hear stories about autonomous vehicles producing four petabytes of data a day.

Our interaction with these cars is not just our personal experience in travelling in them whether they are autonomous or not. The new Audi that we have has a WiFI access point built in so that people can connect. We don’t Bluetooth with the car anymore, but we WiFI to it.

Steve Ardire:

Right, you’re-

Dez Blanchfield:

And as we’re driving, it updates its maps in real time with a built in mobile phone that’s connected to the backbone of the network. I can’t even conceive the amount of data that’s going through that. I just see a little map of me driving around and it refers to a different blue line on the map.

I think you hit on a good point there that the cognitive piece, particularly the AI and machine learning piece, is going to draw out the pieces that we want to see, we want to know about, and it’ll learn the context, it’ll get domain specific knowledge about what I want versus what you want and the time and day that I want it and when I want it, what I want to see when I get out of bed versus what I get at home, whether it’s breakfast or a meeting or lunch and dinner.

Steve Ardire:

It’s spot on. I just wanted to emphasize, with one brief thought is that it’s becoming a manifestation of the best personal assistant, which can predict your needs by knowing you, a personification of you a whole lot better and that’s a good thing.

It’s really helping us in, I don’t like to do. If you can relieve some of the menial tasks, the scheduling, meeting people and then understanding you in terms of your both personal and business needs, that’s a good thing.

Dez Blanchfield:

It goes back to that point that Ginni made, it sort of covered last year around just focusing on the humanities, I think. That is, why are we building this technology in the first place? Why are we investing so heavily? Why is IBM completely pivoting its organisation around these things?

It’s because it’s all about humans. It’s all about people, so the event themes in general, there’s been a lot of them, but we spoke earlier on about the ones that are of particular interest to you.

I’m just going to recap them in point form and throw to you to comment on them. In particular, you mentioned that you loved the concept of empowering people in teams, that you love the delivering on the promise of cognitive computing, in particular, and the synergy between the whole Watson cloud story. How do you see these themes playing out in a day-to-day sense in what you currently do?

Steve Ardire:

Right. Playing out in what respect? I’d clarify that.

Dez Blanchfield:

Sorry. Essentially putting them into practice primarily and then maybe from a consumption point of view. In other words, let’s just talk about the first one. Empowering people in teams? What does it actually mean and how you go about actually making that a reality?

Steve Ardire:

Right. Certain organisations, some of which of course were here at our viewing, like H&R Block, really that’s empowerment to where-

Dez Blanchfield:

That’s a great example.

Steve Ardire:

They’re bringing in the customer to where you’re creating more of a interactive conversation with the H&R Block tax advisor. I thought that was just wonderful to where it’s all about engagement.

Engagement is everything and the better you engage, the more people are typically satisfied. If using machine assist to help with that engagement to create a very positive experience, I see all industries, H&R Block starting to adopt this because it’s customer service is king, and that builds loyalty.

The old way of saying, “Why are people, why do I have so much churn?” these telecommunications companies. Well, you don’t really know your customers, so the burden is on you to know your customers and serve them better and you’ll win.

Dez Blanchfield:

H&R Block was interesting actually. I was quite impressed with the simple, go-to-market message that the gentleman from H&R Block, his names escapes me now, gave us, and that was he showed a screenshot of two monitors.

Now, for the rest of us with two or three monitors doing data science and analytics, we take that for granted, but he sort of explained that the normal practitioner of a tax professional had one monitor and they had put this new innovation of putting a second screen in place, but the salient point was the second screen was for the client.

Steve Ardire:

Exactly and that’s the engagement part. Rather than just sitting there and playing with a smartphone, they said you can put down your smartphone, engage with me, we’ll have a meaningful conversation where we’re using Watson to guide us, but we also want your input and you’ll feel better for that and more confident that this outcome will be positive.

Dez Blanchfield:

I thought it was funny. To them, a second monitor was a major innovation, but what was interesting is that he didn’t miss the most critical piece from my point of view is, “Yeah, okay. We’ve innovated, put a second screen. It’s for the consumer and they’re not going to play with the phone anymore,” but then he immediately went into the key point for me anyway, and that was that the engagement was for the consumer to talk to Watson effectively in a sense.

Steve Ardire:

Correct.

Dez Blanchfield:

Like Watson drew graphs on the screen and started to drill down from the 50 or 60 things that could be bought into this thing from the house and the car and then whatever it might be, fire and risk to these are the three things I think you should really focus on and now that engagement’s not just seeing it on the screen, but actually interacting with that screen and engaging with the system.

Steve Ardire:

That’s right and doing it, though, in a very natural way. That’s the humanity, where we’re not mandating you to learn something. Everybody knows how to communicate whether it be through voice or text input.

We’re making it almost seamless to be able to invite them in without encumbering them or worrying them that, “Oh, because I don’t want to fill out a survey. I want to have a conver … ” Wrapping it in a conversation is really the way to go.

Dez Blanchfield:

An interesting thing that he showed us in the demo they had it even though those screenshots was the tax professionals didn’t have to be retrained in a sense because they just typed the queries in plain English.

Steve Ardire:

Exactly.

Dez Blanchfield:      

What if I did this or how does this affect me? I was really blown away by that. I love the innovation of second monitor. I really get the value proposition of making that available to me as a consumer.

Steve Ardire:

The other point on it, the tax professionals had more enjoyment because these are, a lot of these folks are retired folks, and they’re saying, “I’m having more fun. Rather than me just being a data clerk, so to speak, I’m now forming a relationship with our customers and they’re engaging more,” so that’s a good thing.

Dez Blanchfield:

I think it’s great. I’m just not ready for my tax professional to be a hipster. One of the other themes in particular that I want to just delve into quickly, delivering on the promise of cognitive computing.

That’s a fairly broad brushstroke. If we bring it down to potentially maybe one in particular the startups, which I was very fond of seeing the video that you shared with me because it had an Aussie actor as a voiceover. It was at Meelo.

Steve Ardire:

Cheers.

Dez Blanchfield:      

Yeah. Can you talk about that and particularly how the, I guess the delivering on the promise of cognitive and where that was put into use with that startup?

Steve Ardire:

Absolutely. The company is called, the commercial spin-out is called Soul Machines. They’re a spin-out of the laboratory for Animate Technologies in Auckland, New Zealand, your part of the world, Dez. Mark Sagar, the founder, is just a brilliant, brilliant individual. He’s Academy Award winner, special FX supervisor for King Kong, Spider-Man, Avatar.

He was set up in the lab like four years ago, I believe it was, and he wanted to create a life-like digital avatar that had expressive emotional intelligence. That’s exactly what he delivered on in this commercial spin-out.

He’s still doing wonderful research pushes the envelope, but the tie-in with Watson … When Watson got a hold of this, they were selling to the organisation that you know well, NDIS, which is the, their health care.

This particular effort was for the half a million Australians with disability to create a digital avatar for patient care to relieve the 100% burden on human health care assistance.

The beauty of this, and this is a great use case, is it’s going back to augmented intelligence and the manifestation of not scary AI or just AI for entertainment or games, but useful AI for a purpose to be a digital avatar to help be a friend, to guide, to be there 24/7 conceivably and this is being rolled out this year in Australia.

Dez Blanchfield:

I’m going to get you just to quickly cover the website address because the video on the demo blew me away. I watched it three times in a row. What I saw was the individual in a wheelchair interacting with the AI and the thing that struck me was that that whole human to machine, machine to human interaction again, it was a seamless transition from what I could see in the video.

The person sort of coming up to a screen and interacting and asking questions with a human experience in it. Like it wasn’t talking to an IVR on a phone. I didn’t have to push *1. It made the hairs on my arms stand up when I realised that as your point before around delivering on the promise of cognitive, your startup has already done that, yeah?

Steve Ardire:

I would say so. The whole point of these engaging human-like avatars, Dez, is to create something that has personality and character with embodied cognition. The avatar sees you, hears you and through … It’s not going to be out of the gate through conversational.. effective conversation.. but through text chat, it will be able to communicate with human-like expression. How good is that where-

Dez Blanchfield:

It’s mind-boggling.

Steve Ardire:

This is a subtle or even more dramatic motion response, so it’s really tying upon the emotional intelligence. One of the other quotes I provided is I really think in terms of delivering on the promise of cognitive computing, we can’t forget about emotional intelligence.

I think that’s a key component of future AI going forward because you can’t, as AI becomes more sophisticated, you can’t have human-like intelligence without personality and emotions because people change their behaviour not on information, they change it on emotion.

Dez Blanchfield:

Look. As a parent, I’ve seen the transition from focus on IQ and creating smart humans to EQ.

Steve Ardire:

Bingo.

Dez Blanchfield:

I started writing a paper, and I haven’t finished publishing yet, around this whole topic of where we were focusing on the ability to do “1 + 1 = 2” versus “Is this person happy or sad?”.

And I think that again comes back to where the key valued humanity for these types of technologies or all of technologies in general … The automobile made it easier to get from point A to point B. The airplane makes it even easier.

We get big ships and can move things around us to get from the point A, point B. When we think about technology, it needs to essentially focus in on and gravitate around humanities.

Steve Ardire:

Totally. It comes back to that was probably my biggest from Ginni’s was that point. Come back to humanity and in many respects, AI is a huge driver to fostering that, to be able to cut through special interests, moneyed interests. This is exciting for us.

Dez Blanchfield:      

Absolutely, and as you said, particularly in my backyard in Australia, the NDIA, the National Disability Insurance Agency who run the national disability insurance scheme, as you said, they’ve got a pilot programme of 500,000 people who either have a mental disability or a physical disability. Look, even if you’ve had a fight with a forklift and lost your left arm, typing into a computer is not easy.

You’re hand pecking. Better interact with a familiar face and a familiar voice in a vocab and in a language that’s common to you. It completely removes that whole challenge. Last year at World Watson, I’m sure you remember this, we got to interact with Pepper, the robot.

Steve Ardire:

Right. That’s right.

Dez Blanchfield:

There were a lot of comments that I thought were unfairly negative around the limitations of Pepper because people were asking things like, “Do you love me?” whatever, but it turns out the Hilton network are working with IBM and I think it’s EY, to roll out tens of thousands of these around their hotels because they don’t want you to have a conversation as, “How do I feel?” But it’s, “Where’s the nearest bathroom? How do I get an extra towel?”.

But again, it comes back to the human component of how does technology help me, right? I think we’re at the exciting cusp of when we might actually see it, as you said, not just delivering on a cognitive piece in your interest, but in technology delivering on a promise as a whole to humanity.

Steve Ardire:

Exactly. Between the, in the financial services that I spoke about and health care, well this is a health care example, but it’s bringing in that emotional intelligence with the humanity, which I think is just wonderful.

Dez Blanchfield:

Absolutely. Let’s face it, if you don’t have health and you don’t have money, you can’t look after yourself or anyone else.

Steve Ardire:

Indeed.

Dez Blanchfield:

Mindful of time, because we’re aiming to keep this one to half an hour, couple of last quick ones and I’m going to ask you for kind of your medium-term view of what’s over the horizon.

We’ve heard about the Watson data platform. We’ve heard about Watson machine learning. In particular, data management and data governance, which is becoming more and more of a challenge.

There was a great line yesterday where the reality check is governance and regulatory compliance is only going to increase and get more complex. It’s not going to decrease. I have a question to throw to in a minute there.

Particularly in the integration of how to open Spark and the data science experience, with data governance and build around the team collaboration challenges and the promise of cognitive computing in particular, you gave me the quote of, “AI and the future of work.” Data governance in particular I see is a really big challenge because when we put everything into a database we can control it.

When we put it into a data lake, we lose a little bit of control, but we gain flexibility. When we go from the data lake to bursting to a machine language engines to cognitive engines like Watson, we’re sort of taking subsets of the data or metadata out into another place and then bringing it back.

I’d really love to get your insights and views and opinions on what that means to data governance and how we kind of even just grasp the concept of allowing some of our data to go to another place, whether it’s to a public cloud offering of machine learning on the likes of IBM’s Bluemix and data Watson platform.

What does that look like? How do we explain to people what that means to apply controls in data governance to data we’re allowing to move to places we really haven’t experienced before and probably don’t have design patents for.

Steve Ardire:

Right, and there was an interesting slide, I forget from who that presented at one of the general sessions, in terms of the hybrid cloud, how that’s like dominating going forward. I really think it’s as if we take that metric of 80% hybrid cloud, I think that’s really going to be data governance for that rather than private and public, you’re depending on the cloud vendor. It’s going to … Some combinatorial instantiation of that and I think Blockchain is clearly dialed right in there.

Dez Blanchfield:

It’s a game-changer.

Steve Ardire:

It’s a game-changer. Blockchain will be embedded into it. I’m not a Blockchain expert. I keep up-to-date as much as possible, but you know that’s fundamentally just like you do smart contracts, why can’t you do smart data governance is my question.

Dez Blanchfield:

Well, in fact, yeah. It’s interesting you mention smart contracts. The concept of the smart contract really it’s just a control mechanism.

Steve Ardire:

Correct.

Dez Blanchfield:

And it’s being logged and in fact my view of this on what you touched on that is that Blockchain’s probably a technology like cognitive computing that people won’t necessarily interact directly with. It’ll just be built into the engine-

Steve Ardire:

It’s built into the engine. Absolutely.

Dez Blanchfield:

I don’t really want to know about [TCIP 00:38:23] or DNS and domain name look-ups. It’s just there. I type in the URL. It converts it to a number, so no. I think you’ve hit on a really good point there.

Steve Ardire:

The other corollary to that, Dez, is again I really think that like Meelo is an AI cognitive service infrastructure, so basically the whole point of that is it’s an overlay, a cognitive fabric overlay.

It’s not requiring you to, it’s assuming that the data lives where it is. We’re just overlaying it and then we can tease out the things like temporal contextual progression to be able to better understand the context of actions and content.

To me, it’s kind of a new form of data governance on the fly that really kind of dovetails with, “Oh, wow. That means I don’t have to do data mining and analysis. It’s like embedded like Blockchain into the process.” I think is really exciting stuff.

Dez Blanchfield:

I think it drives us to a point where effectively data’s going to have to be self-describing in many ways.

Steve Ardire:

Correct.

Dez Blanchfield:

We had things like JSON, XML and [inaudible 00:39:29] files who when you look at the data, it describes itself. We did that from the schema point of view so we could have schema-free business intelligence [crosstalk 00:39:38]

Steve Ardire:

To that point, I just, talk about the beauty of coming to these conferences. I was wearing a Kimera dot AI shirt yesterday when before we went out to do our contribution to the feeding thing and stumbled a wine bar, this guy said, “Kimera. I know that thing Kimera.” I said, “Yeah, we were at AI World last year.” He says, “Well, let me show you what we’re doing”.

As far as the data self-describing, he showed me a hypergraph architecture. They are working with IBM and I was very, very impressed. This is where you can … Self-describing where they can take entities and concepts and understand the relationships and semantic bindings between it. Very impressive stuff.

Dez Blanchfield:

That’s a brave new world in itself, right?

Steve Ardire:

Yes, it is.

Dez Blanchfield:

Wow. Self-describing data.

Steve Ardire:

Yes.

Dez Blanchfield:      

To wrap up quickly, I’d like to throw at you my favourite pun. Watson the horizon for the rest of 2017? I think trying to go beyond that is just-

Steve Ardire:

Yes.

Dez Blanchfield:

It’s a very difficult challenge, but for the rest of this year with what we’ve seen so far, off the top of your head, what do you think is on the horizon for the next sort of six to 12 months for not just folk here, but outside of this, what we’ve seen around the Watson data platform, the Bluemix cloud, cognitive, data first?

Steve Ardire:

I think we’re seeing deep learning really be one of the dominant full-on … I’m very excited about transfer learning. Transfer learning, which all the big players are doing right now as more.. All the big players are open-sourcing their engines out there, but transfer learning is where you can train your data and supervise and then take that and apply it to a different domain because remember we heard here, you want to go deep and wide in terms of the domain, so if you have methods like transfer learning, you can actually stand up these embedded machine intelligence systems a whole lot faster.

One of the things I’m looking for strong improvement is unsupervised learning. Most of the work today is supervised, like I just described there, but unsupervised learning is where machines can infer what they don’t know and that’s exciting because to really build intelligent systems, that’s where the algorithm kind of learns on its own.

There are some negative sides. There’s been some posts regarding, “Well, if you factor in bias into the machine learning algorithm, you’re going to get biased algorithms.” There’s been studies, like in a beauty contest, that particular algorithm chose mostly white Caucasians as the winners. That’s an example.

Dez Blanchfield:

Right, right.

Steve Ardire:

Exactly.

Dez Blanchfield:      

Pretty serious consequences.

Steve Ardire:

Serious consequences, but I think that the unsupervised learning for really complex problem sets where the training is kind of like one-dimensional. You can kind of go multi-dimensional, so I’m very, very keen, very keen on that effort for progression in 2017.

One of my favourite things that I say is that when you’re getting into, and we’re not there now, if you can actually crack transitive reasoning, and this is what humans do naturally, is that they can maintain context from one subject to the next, that is really killer and that’s where you really can start to manifest intelligent AI that can interact with you in a very natural, fluid way.

Dez Blanchfield:      

What a fantastic point to leave on, transfer of reasoning because I think that when we interact with our current smartphones, we talk to the phone, not with the phone-

Steve Ardire:

Correct.

Dez Blanchfield:

I think that’s great. Transfer of reasoning. I think we need to do a blog on that.

Steve Ardire:

Transitive.

Dez Blanchfield:

Transitive reasoning. There you go.

Steve Ardire:

Transitive.

Dez Blanchfield:      

Sorry, thanks for correcting me.

Steve Ardire:

It was transfer learning and transitive reasoning.

Dez Blanchfield:      

Got you. Perfect. We’re going to have to blog those. Steve Ardire, it’s been a real pleasure chatting with you and I’ve had a fantastic time hanging out with you and the team here at IBM’s InterConnect 2017 in Las Vegas.

Thanks for your time and look forward to doing it again soon.

Steve Ardire:

Thank you, Dez. It’s been wonderful and I agree with your comment. It’s been a wonderful event.

Dez Blanchfield:      

We have. Thank you very much, Steve.

Dez Blanchfield

Dez Blanchfield is a strategic leader in business & digital transformation, with three decades of global experience in Business and the Information Technology & Telecommunications industry, developing strategy and implementing business initiatives. He works with key industry sectors such as Federal & State Government, Defence, Banking & Finance, Airports & Aviation, Health, Transport, Telecommunications, Energy and Utilities, Mobile Digital Media and Advertising, and Cyber Security. His focus is driving outcomes for organisations by leveraging Digital Disruption, Digital Transformation, Cloud Computing, Big Data & Analytics, Machine Intelligence, Internet of Things, DevOps Integration, Automation & Orchestration, App Containerisation & Micro Services, Webscale Infrastructure, and High Performance Computing. Be sure to follow Dez on LinkedIn ( http://linkedin.com/in/dezblanchfield ) and Twitter ( http://twitter.com/dez_blanchfield ).

You may also like...