This is the Nielsen Norman Group UX Podcast.
I'm Therese Fessenden.
It's fair to say things are a bit tumultuous in the user experience world right now.
There are lots of layoffs rocking big tech and threats of artificial intelligence plagiarizing
us or making us redundant.
While it might feel unprecedented to younger generations, this isn't the first time there's
been a major shakeup in the tech industry.
In 1996, there was a major boom in tech known as the dot-com boom, when thousands of websites
practically erupted into existence.
Except it didn't keep that name.
The dot-com boom later became the dot-com bubble, and then in 2001, the dot-com bust.
Two million people were laid off in 2001.
After hundreds of publicly traded companies went bankrupt.
Today you'll hear my chat with someone who not only lived through this, but all the while
continued advocating for more user-centered websites.
That person is one of NNG's co-founders, sometimes referred to as the guru of usability,
Dr. Jacob Nielsen.
In our first episode, Jacob gave us an overview of UX and usability.
In this episode, we do a deeper dive into how he got into the field, when frankly the
field didn't exist, and what's changed over time.
Whether or not all this talk about generative artificial intelligence is just a phase or
truly the next chapter of UX work.
When I got into high school, I actually started using computers, which I think was because
it was so boring for me to be in high school, but the computer saved me because it allowed
me to do interesting things at a more advanced level than was at my level, I guess.
Then I actually started studying computer science because I've been so thrilled with
using computers.
I really was, I guess, kind of truant from a lot of the official classes and to spend
all the time with the computer.
That gave me that feeling that computers could be pleasant and easy to use.
Then I started studying computer science, and that was a big letdown because then we
were given this huge mainframe that was in the basement of the computer science department,
and it had a terrible, terrible user interface.
It was very, very unpleasant to use.
Compared to that much less powerful computer I'd used in high school, it couldn't really
do anything.
We just made very simple little games on it and things.
That made me think, well, I knew it could be different.
That was a crucial point.
Then I think another influence in me was my parents, both of whom were psychologists.
That gave me very early experiences, kinds of testing as well.
I think those experiences meant that I thought that, well, we can do better than just have
to accept the way computers are.
We can run experiments with humans.
We can learn how humans work.
We can figure out how to do it.
That's actually what I did when I got into graduate school.
What year was that when you were getting your PhD?
It was early 1980s, I mean, mid 1980s, yes.
I'd worked on it for like four or five years.
I was a university professor for a few years, but then I also kind of figured out that's
not really where I want to be.
That's when I changed and got a job at Bell Communications Research.
I'm just thinking you were basically a usability expert during the dot-com boom.
I would love to know how that was like.
What was it like to be that expert during this time?
Well, I mean, in many ways it was a happy time because... It's fun to say because the
user interfaces, the websites were so bad.
Because they were so bad, it meant that it was very easy to improve them and to affect
major changes in the business.
Today, websites are still not perfect, but they're much better and it takes much more
time to dig into a website and really figure out how to make it better.
But yeah, it was in many ways very happy.
And it was also this notion of when you have new ground to explore, which was the web.
I mean, the web was growing at this enormous speed of several hundred thousand percent
per year.
It was really, it was boom time, enormous boom times.
When you're going into an experience established field, you have to kind of work at small things
instead.
So in the beginning, we were discovering all these really big things about like how people
do search or even that they do search because search before the web was very terrible and
really only for our librarians or the specialists.
And so yeah, people use search, people are extremely impatient.
All those discoveries that are fundamental to us today, we made.
And a lot of them, I remember actually one of the, I think the very first report I wrote
on web usability, it was just based on having tested, I believe like four different websites
with just a few users.
And already it was, whoa, this list of enormously groundbreaking insights.
And then of course we kept doing more and more after that.
But what I should say is, Doc famously prepared mine.
I forget who said that, but that's kind of one of the famous sayings.
And so I was prepared because I'd done many years of previous work on usability on how
to do user studies, how to learn about how people use computers.
And I'd also done several years work on hypertext, which is the foundation of the web, which
is how you click on one thing and you get another thing.
And so I was already an expert on these two things that came together and became the web
user interfaces.
And so that of course gave me that kind of like headstart on when the websites were there
to actually study and actually discover these new things.
But I would say it was extremely exciting.
It was also a period of like 80, 90 hour work weeks because it just boom, it just came coming.
There were so many things happening.
Yeah, it sounds really busy and also very lucky, but as you were saying, like luck favors
the prepared.
So you ended up having just the right skillset at just the right time.
And what comes to mind now is I'm thinking of how UX has changed, right?
In many ways you've said it's improved significantly.
And obviously there's a lot of really, really new stuff that we're going to talk about shortly.
But what would you say is like the biggest difference that you've seen in the UX profession
between then and now?
Well, I mean, it is certainly much more professional or it is a profession whereas in the old days
it was just a handful of people who were playing around doing this stuff.
What we do is in some sense not that different because we still use these basic methods of
observing people, giving them real tasks to do, not just show you, here's a demo, don't
you like my design how great it is?
Those methods are exactly the same as they've always been.
So I feel like that also is, by the way, the robustness of the methods is very important
I believe because that's why it's worthwhile to learn them now because you can still use
them in 20 years.
I mean, I don't know what the science people will be working on in 20 years, but they can
use the same methods to do studies of them.
And so that I feel is a very constant in our work is the way that we study people.
But the other change I want to say about the UX profession is just more.
It's bigger.
There's more people, more users, I mean, really there are billions of people using computers
now, which I was not in the past, which makes it so much more important for humanity, for
the world to make good user interfaces and more people doing UX and more user interfaces
that need to be better and all those things.
So it's just scaled up in magnitude enormously.
But if you go back to like, I'm one person working on one design for this set of target
audience, then it's actually kind of quite the same.
Yeah, right.
A lot of the same principles, a lot of the same processes in order to get to that outcome.
Now it's funny you mentioned things growing at scale.
I think what's on a lot of people's minds at the moment is AI and how generative artificial
intelligence can really revolutionize the way that we work.
So in a previous episode, we had Susan Farrell, former NNGer who came to talk to us about
what AI is, some of the risks associated with AI systems.
So obviously right now there's a frenzy of interest and I'm curious what your thoughts
are.
Do you think generative AI is a phase or is this something that's going to be a more substantial
force in our future?
I do think it's going to be a substantial force and one more thing that computers can
do, there's already been advances in AI in other types of things like recognition.
So for example, we are now having subtitles when we do our courses online and we're just
going to start having translated subtitles, which again is all done by AI and that's a
major advantage to again for the expansion of UX, which is so much more worldwide.
Those are kind of have been coming along for many years and are just gradually getting
better and better and better.
Whereas I think the generative that the computer can come up with something new is not something
we in the past maybe thought so much about, but I feel it's an extremely useful phenomenon.
So I do think it's something that will be embedded in a lot of applications in a lot
of places around the world.
I honestly think that we should be more optimistic than that because of course right now they're
not perfect.
They'll probably never be perfect, but one thing we know about AI is that it does get
better and better and better.
And so right now if you take something like this chat TPT and come up with text that sounds
very credible, but it's actually wrong.
And so maybe in the future it will be less wrong.
But I also think that that's not the correct way of thinking about it because I would rather
have these things be used in kind of collaboration with humans.
And one thing you can do is if you get the computer to generate, let's say three or four
explanations of something, then you can look at them and say, which one of them is actually
the correct or is the better one?
And then you can use that.
Or if you have pictures, you have to generate three or four pictures, which one is the best?
And then you run with that.
So the function of this an art director or an editor or something like that becomes more
important and the function of the photographer or somebody who can string sentences together
so that they flow well become less important.
But that's just like so many other tools we've had in history.
But when I was in high school, the big debate was calculators.
Now, should high school kids be allowed to have calculators because then they would cheat
on the exams and they wouldn't learn how to calculate and so forth.
And you know, honestly, today, you know, if I need to add up to 10 digit numbers, I don't
say do it by hand, I mean, Excel do it.
So it's not important to teach people these more primitive steps that can be done by something
by a computer.
But what is important is to teach people how to use those tools.
And so that I think is what we're going to go.
And I have no worries about nothing people often bring up is all the unemployment, but
because computers will replace us, but all experience from history shows that this is
not what happens.
This is what's called a fixed work fallacy, which is that there's only a certain amount
of work that needs to be done in the world.
And if some of that work is done by the computer, then the people who do it will be unemployed.
But you can go back to even like the stone age when agriculture was invented.
So farmers can create much more food than hunters.
And so does that mean all the hunters were unemployed?
No, they just became farmers instead.
Now what is an issue is, of course, and when it happens fast, then people need to be reskilled
or retrained because they may have been used to doing one of those jobs.
And the biggest scheme of things, I don't feel any unemployment is going to be an issue.
I think it's actually the opposite, which is like lack of workforce, that's an issue
that we need.
We need more, more people, not fewer.
Yeah, what I think is really interesting is what you said about like the types of skills
needed are going to change.
But one that comes to mind is like an AI that generates different UIs, right?
Or if you ask an image generator to come up with an image of a mobile application that
allows you to grocery shop or something like that, it'll create a little interface for
you, but there will be tons of issues with it.
If you are seasoned enough in the field of UX, you almost see it right away.
You're like, wow, that contrast is pretty low, or these buttons need to be a little
bit bigger, or various other different usability issues.
So what I think might change isn't so much that, oh, UX designers are going to be out
of business.
No, if anything, the work of design is going to be much more focused on refining designs
and making them more appropriate, more specific to whatever it is you're really designing.
And in a way, like using it like a calculator.
And then choosing, and again, going back to the point about math, what's important about
math is not to be able to add up too tentative numbers, but try to understand what it means
that if one number is much bigger than another number, what's the implication of that for
your business and things like that.
And similarly with user experience, what's important is to understand the humans and
their tasks, and not necessarily to create new lines that are perfectly aligned, because
the computer can do that for you.
And so again, I feel that this notion of having to generate multiple things, and then you
choose the right one.
I think that's very powerful.
I'd actually like to tell you a little story from back when I worked at the phone company
at Bell Communications Research.
This is now 30 years ago.
So 30 years ago, we came up with a principle we called N of 2N, which it was, again, not
a marketing person named that.
But the idea was that if you need N things, you actually produce 2N, you produce twice
as many, and then you give it to the person, and they select which ones they want.
And we actually, even back then, this was optimal compared to human assignments.
So we were actually testing it out on a problem where you need to assign, I mean, we had to
assign scientific papers to reviewers, that was our task.
But so it's normally done by having a human expert say, well, Professor X would be a good
person to review this.
But it turned out that that's not such a great thing because first of all, all these different
professors, they're all kind of famous for what they used to do.
So they all got assigned to review papers and stuff they worked on 10 years ago, which
they're not so interested in anymore, but which they're famous for.
And then secondly, you get obscure topics and you have no idea who's a good expert on
that.
But the computer can figure all that out.
And we actually did, we got the feedback from all these experts after we did the experiment
on whether they thought that they got better outcomes.
So the computer assignments were actually better than the human assigned.
But with this caveat that we gave people the choice of twice as many things as what they
actually had to do.
So we called n of 2n.
So just by doubling, so the computer could not be accurate.
Now it was also 30 years ago, so maybe it could be better today.
But it could not be accurate, but it could be accurate enough that if you gave people
some choice between the things that the computer gave them, then they were actually more happy
than what they could get from the best available human expert to produce.
Wow.
That's really fascinating to hear that.
And I think you're onto something with the good enough results.
I don't think people are necessarily seeking perfect results when using these different
generative AIs, but looking for something that is good enough.
And then I think the other aspect, which is interesting, is still giving people this feeling
of choice, which is if you were to double the options or give people multiple options,
people still have control over the end outcome, whatever it is they're producing.
Or I don't know if you're using chat GPT to make a haiku card or something that you generate
a few haikus and then maybe one of them, or maybe the three of them, you can pick a few
good things out of all of them.
And then at the end of it, you have a superior result than what would have been the case
if you started with nothing.
So it definitely seems like it'll be a good way for people to start projects, or at least
to start coming up with solutions for various different ideas.
So it's very exciting.
Now thinking of the future, how are you feeling overall about the future of UX?
What keeps you inspired to keep doing what you're doing, even all these years later?
Well, I mean, I think there are many, many things, but what one of them is that it is
this true growth area.
So it's getting bigger and bigger and bigger.
And we only really scratched the surface yet of what we really can do to contribute to
the world, because there's still the vast majority of user interfaces are still bad.
They're not as terrible as they were 40 years ago.
So I think, I mean, the good way of looking at it is the glass is half full and half empty.
So what we do works.
We actually deliver results.
We actually improve things for humans.
So that's proven.
I mean, it is in fact much better than it used to be because of us.
But at the same time, it's not nearly where it should be.
And that's why I can kind of see that we will be able to bridge that gap and create a world
of technology that's so much more suited for humans than what it is even now.
And I feel like that's something we can do.
So there's enormous growth potential and use need for so many more UX people.
I don't think the problem is unemployment.
I think the problem is too few.
So we need more.
We need more.
And this is one of the things I am very motivated to do is help create those more experts by
broadcasting, like this podcast, I guess, but in various media forms, distribute expertise
about how to do it better so that it will in fact be done better.
I feel that's very exciting.
And there's also thing that's also other exciting is that there's always also just new topics
that come out.
Now we just talked about AI, or 20 years ago it was mobile and 20 years from now it'll
be some other new thing.
But there's always something new and sad to say, one of the lessons from history is that
when something new comes out, it's usually done wrong and that creates a great opportunity
for us.
That's a positive way of looking at it.
But the sad thing is that it tends to be when some new technology comes out.
It is more driven by, you'll say, wow, we can build this fancy new thing.
And that's great because that's their motivation.
So we need people to invent all the new stuff.
But at the same time, they tend to invent things that are useless because they don't
have that second side of the equation, which is the people.
And that's our side of the equation.
So we match the two sides up and then together we create a better world.
Yeah.
And actually on the topic of building new things and more things and how there's no
shortage of UX work, I actually wanted to pick your brain about another topic.
In thinking about the pressure for a lot of companies to build more, either new features,
more features, but not always the most useful ones, as opposed to maybe revisiting existing
infrastructures or existing features that maybe aren't working as well as they should,
like what comes to mind for me are some of the major tech infrastructure failures over
the last few weeks, especially in the aviation space.
So I was wondering if you had any thoughts on how companies can or should be balancing
this chase of the new and pushing the envelope with innovation, but at the same time reconciling
the UX debt that they may be having.
Stability is a feature in its own right, and it's actually getting to be more and more
efficient.
Like in the really old days when there were very few features implemented, the added value
of another extra feature was actually quite high and so therefore worth doing.
Today, computers have thousands of features.
Any software you might use has so many features that it's too much.
And so the incremental value of one more is actually very low.
But at the same time, the incremental value of fewer crashes, a few times it goes wrong.
That's huge because people waste enormous amounts of time.
Now, it's of course particularly bad when something like the air traffic system is going
down or something like that, where you have a lot of people being stranded and wasting
hours or days potentially.
That's extra bad.
But even sort of smaller things like on your own computer that you click something and
something wrong happens, most people are not able to figure out what went wrong.
And actually, I cannot figure out what went wrong most of the time when I have a PhD in
computer science and I still can't do it.
And it's just the complexity is so high now that when things go wrong, we cannot identify
the problem.
That means that you spend a lot of time hunting down, testing five, 10, 20 different hypotheses
that are wrong.
They cannot do anything.
And so you waste a lot of time on that.
And people build erroneous mental models.
They sort of fear that, oh, if I click here in a certain way, the computer blows up.
And that was not the reason it blew up.
There's some other underlying reason that we don't know because it's unstable software.
And so you are really degrading the user experience, which means that of all the existing features,
their value has been lowered by some substantial amount by the instability of the software.
So I feel like making it just work, I mean, just solidity, like work, work, work all the
time without fail.
That would be so enormously valuable, much more than any one or two or 10 new features
that people might implement instead.
And security is, of course, another area of that where there's endless break-ins and crackers
that get in various your accounts and is enormous complication.
That's another area where making it work better would have enormous value for everybody, basically.
Absolutely.
I couldn't have said it better.
And it's funny how you mentioned the mental models and how I know plenty of people of
all ages who are terrified of using certain computers, or they've just basically deemed
themselves, oh, I'm not a computer person.
I must not be good at this, when in fact, maybe it's the design teams who put all the
things together that they had worked on, which unfortunately have tainted the perception.
Oh, completely.
Well, that's, of course, I mean, there's a little bit of ideology to the user experience
field.
And one of our key ideologies is it's not the user's fault.
It's always the designer's fault.
I mean, this is a little bit overstating it because it's not the designer's fault.
It could be the programmer's fault or somebody else, but it's the system's fault anyway.
It's not the person, the user's fault.
So we have to take responsibility.
The collective, we have to take responsibility.
Both of the design, like we actually put these words on the screen or buttons on the screen.
That's the surface user interface we're designing.
But again, we want to talk about total user experience and total user experience includes
the implementation quality.
So every single engineer and program or developer, they are part of that user experience.
And if it fails, when it breaks, when it crashes or it doesn't do things right, any kind of
these type of problems, it also becomes a user experience problem.
And so again, it's the fault, it's the collective fault of all these people who contributed
to building the system.
It's not the user's fault.
I mean, the user is like they are.
That's one of the other ideologies we have is that we have to accept the world the way
it is, not sort of some idealized, oh, it would never be wonderful if everybody was
like super intelligent and they would like read every single word of instructions carefully
before using our software.
But we have known for like probably about 40 years that users don't read instructions
and we cannot change people and people are the way they are.
And so we have to change the system to work with the way people are.
And that's the kind of the foundation for everything we do.
It really is.
And again, I think speaks to the job security that we will continue to have unfortunately
and unfortunately.
Exactly.
So actually related to that, if thinking about those folks who are just getting started with
UX, I'm sure they kind of feel a bit overwhelmed, often thinking about all the things that they
may need to be thinking about and considering, do you have any advice for those who are just
starting?
Well, I think the best advice is really to do it because UX is a field that's more based
on experience and it's based on theory.
I mean, certainly there is some theory, there are some methods, for example, for how you
do usability testing or any of the methods we have, there's like some principles for
all of those and there are design principles and so forth.
So there's a lot of, and not a lot, there's some theory that's definitely good to have
because if you don't have that, you're just like flailing and you don't know what you're
doing.
So absolutely learn that.
But then after that, most of the quality of good UX works come from experience and having
tried and having done it.
And so that's therefore, doing it is really the only advice.
And of course, the first time you do something, it's not going to be perfect, you just have
to accept that and it's going to be better the next time and even better the third time
and also do that kind of self-reflection.
That's another thing that's very important.
I think a lot of people don't do, like for example, let's say you do usability testing,
you could record the video of that and then watch it with the purpose of reviewing your
own performance, not review what the user is doing, which is what we usually do, but
review yourself or even sit with a colleague and you can review each other and give these
other good pointers on how you can improve.
That I think has done much too little, but that builds a reflection of professional performance.
That's also true for students.
I think they are often a little bit held back by a lot of theory and by being in places
where they don't do enough kind of hands-on or practical work.
But then I would say for my students, my best advice is get internships.
So go out there and spend the summer really hands-on, diving deep into a project and learn
from that.
Yeah.
Oh, definitely.
If I were to think about the difference between my college education and my first job, I think
my first job, I mean, I absolutely used what I learned in my undergraduate degree program,
but it didn't really make sense until I was there, until I could actually talk to people,
work through problems.
So I a hundred percent agree with that.
Now, for folks who maybe are more seasoned, maybe they're facing this future, worried
if they're going to be able to reskill quickly enough, do you have advice for those who are
more seasoned, looking forward into the future of UX and the future of their careers?
Yeah.
Well, I think if you are people who have done UX for a while, the truth is that most of
these things come back, even in a new technology, like a lot of, for example, of what we learned
in doing mainframe.
I also worked at IBM back in the old days and a lot of the lessons we had from doing
mainframe user interfaces came back on the early web because the web used to be not highly
dynamic, but rather you would fill in a form, click submit, and then the computer would
go away and come back with the result, which was very similar to mainframe interactions.
And so a lot of what we knew from a very primitive technology was being transferred onto this
news hyped up web thing.
And I think it's the same for many other things, that you're doing about the design patterns
and, again, particularly the human usage patterns and the methodologies for how to do the work.
Those all transfer into the new things.
And then, of course, you've got to change the vocabulary, speak the language.
We always have to speak the user's language, which is one of our famous basic heuristics,
speak the user's language.
And including when you're working in a team, you speak the language of that team, which
tends to be changing, which is highly annoying to me, but it is what it is.
And so you've got to adapt to that.
Otherwise you're going to seem like somebody who's out of touch.
But the under phenomena, the real thing underneath are honestly much the same and they manifested
slightly different ways.
You've got to understand that, of course, and they may be called slightly different
things.
But you can also reuse experience.
Experience is really key in UX, and it can be reapplied to a new context quite well.
Yeah.
It seems like the way forward is being able to adapt, both adapting your skills, yes,
sure, but also remembering that the knowledge you have is adaptable and that there are going
to be many different ways that we can make an experience, capitalize on some of the things
that worked really well in the past and make future experiences even better.
Exactly.
I mean, then we also see the same mistakes being made again and again, and in many ways
history does repeat itself, but I think, and now I forget who said that, but history doesn't
repeat itself identically the same.
There's always some tweaks and differences, and those are of course important to be aware
of.
That I think is key, right?
But basically a lot of things do repeat.
Yeah.
What was it?
I got to look it up later, but I think it was something along the lines of history doesn't
repeat itself, but it rhymes, something like that, but yes, that's very true.
Well, Jacob, it's been a pleasure having you on our show.
If you wanted to point people to any resources, if they wanted to learn a little bit more,
where could you point people to?
Well, I would of course point them to the Nielsen Normal Group website for thousands
of articles, but now that we talk about some of these things, I would particularly actually
point to a few of the keynote presentations I've done in the last few years that are on
our YouTube channel.
And so the video channel are recordings of some of these keynotes where I talk about
some of these things about the UX in 25 years or what I think are going to be some of the
future trends.
And so those keynotes can also, I think, might be worth watching for sure.
Absolutely.
Yeah.
They're all great, especially if you like videos, you like to see some visual references,
I highly recommend them.
But anyway, Jacob, thank you so much for your time today.
I've had a great time and I'm looking forward to chatting some more later on this year.
Yeah.
Thank you, Therese.
It was great.
Thank you.
That was Dr. Jacob Nielsen.
You can find more information about him and the work that we do at our website.
You'll also find information about our upcoming events.
We have virtual events in March, April, and May.
And if you register for any of the courses presented at these events, you're also invited
to Dr. Jacob Nielsen's keynote address at that event.
On that note, we also have a free weekly email newsletter, which features both our upcoming
events and the latest research that we've published.
So if you want to get notified about these, or just want to learn more in general about
the work that we do, go to nngroup.com.
That's N N G R O U P dot com.
Finally, if you like this show and you want to show your support, please leave a rating
and subscribe on the podcast platform of your choice.
This allows us to get the word out about this show and the work that we do.
This show was co-produced by me and Samita Takala.
We're really excited to have Samita as part of our growing production team.
All sound editing and audio post-production is by Jonas Zellner.
Music is by Tiny Music and Shimmer.
Thanks for listening to today's episode.
Until next time, remember, keep it simple.