Is Aggressive Marketing Influencing UX of AI Agents?

After his keynote at the Las Vegas UX conference,
an attendee asked Jakob Nielsen whether aggressive marketing is influencing the user experience
of intelligent agents, like Alexa. [Jakob]: Well, that’s a good question, but
I actually don’t think so. I really think that it is– well, they’ve
been over-hyped, for sure. That, you could say is a fault of the marketing. On the other hand, it’s also definitely the
fault of the usability in that people don’t turn to them for the needs they actually have. So people can generate a long list of things, “I’d like the thing to do this, I’d like the thing to do that.” But they don’t turn to the thing for that,
because their experience has been bitterly negative when they have tried in the past
various things. So, what we found in researching this is that
users tend to stay with a very narrow, limited set of tasks, despite the advertisements marketing is,
“Man, Alexa comes with thousands of so-called ‘skills.'” It can do that hypothetically– technologically, it
can do that. But, from the people perspective, it cannot
do that, because people try a few times, and they fail, and they don’t want to do it again. If you promote something as being “it can
do all these great things,” and people try a few of those things and it fails them–
then they’re not going to try very much more. You really set yourself up to fail in the
long run, because what’s called resampling, which is– again, people tried it, they sampled
the thing already, and now, next year you say, “Let’s better try again.” Well, people have bad experiences — they’ve
been burned and they don’t want to do it again if they have a negative experience. It’s — in the long run, which is longitudinal
user experience — which is rarely considered. But if you think about that long-term experience,
it’s actually quite problematic to over-hype something and make people get the bad experience of
it having no usability, then they’re not going to try again.