The Rise of Conversational AI, Explained
Key insights of today’s newsletter:
Since personal computing’s inception in the 80s, we’ve shifted from command-line interfaces to more intuitive graphical user interfaces.
The advent of conversational AI has reversed the ‘locus of control’, enabling computers to understand and respond in human language.
The impact of this shift is profound, it’s not just a technological revolution; it’s a societal one, as AI is poised to reshaping the future of work and digital life.
↓ Go deeper (10 min read)
Last week, I gave a Masterclass at Future Proof Lab by
. She helps businesses and individuals thrive in the complex and uncertain future world of work.My goal was to help people who haven’t been completely immersed in the conversational AI space to understand what’s been going on. Why are we seeing AI assistants everywhere? Why now? And how should we think about these new technologies going forward?
I covered quite some ground, so read along or sign up to watch the full recording here.
Why now?
Text-to-text, text-to-code, text-to-image, text-to-video. It’s clear that the future is conversational — but how did we get here?
Since the invention of personal computing in the 80s, there has been a gradual shift in the way we interact with technology. In the beginning, we had to talk to computers in their language, directly via the command line. Advanced programming languages helped us do that more efficiently.
Over time, we developed graphical user interfaces and we witnessed the birth of UX design. Touchscreens introduced new modes of interaction, like swiping, pinching, and tapping. With these new affordances also grew the need for experts that could design for these new experiences, better known as UX designers.
The latest shift has been towards conversational AI, which has allowed us to talk to computers in our own language. We reversed the so-called ‘locus of control’. Instead of having to learn how to talk to computers, we teached computers how to talk to us (hence the name of this newsletter). We came full circle.
The human and the artificial brain
This new way of interacting with technology required us rethink how we design for these interfaces. It brought about new skillsets and workflows, and the role of the conversation designer emerged.
Conversation design is the art of facilitating meaningful exchange between humans and computers. It’s a challenge, since we don’t speak the same language and there will always be a gap in understanding. Computers need structured data, like intents, entities and variables, whereas humans understand language intuitively. We are master conversationalists.
Research has shown that we respond very positively to computers that are good at mimicking human conversation. The goal is never to pass the system off as a human, but to make conversations flow as naturally as possible, akin to speaking with another human.
Conventional chatbots vs. LLMs
The launch of ChatGPT, roughly one year ago, turned the world on its head. It showcased a radically new approach to building chatbots.
Instead of designing fixed answers to fixed questions, large language models leveraged a new, revolutionary architecture that relies on predicting the next word in a sequence.
The conventional way of building chatbots is slow and requires a lot of manual labor, but the new way is not easier per se. The idea that you can simply plug-in a large language model and all your business problem will be solved is deeply misguided.
Because these models are so hard to control and their output is inherently non-deterministic, deploying them into an enterprise setting is not without risk. It’s one of the core reasons for why adoption is slow, even though the willingness to innovate is there.
The impact on work and society
It would be a mistake to think of conversational AI as just as a technological revolution; it’s a societal one as well.
Modern phenomena like AI companions and lifelike voice clones, which I’ve covered in previous articles, are poised to changed society, whether we like it or not.
In the workplace, it’s shaking things up as well. It has led to at least two new breeds of workers: centaurs and secret cyborgs.
Centaurs are the people who see AI as an add-on, but are not necessarily over-reliant on it in their day-to-day work. Secret cyborgs are the people in your organization who use AI without others knowing about it. Paradoxically, at least some of the value of AI comes from people not knowing you are using it. AI-generated content that passes as human-written is great, but only if people think it is coming from an actual human.
Nowadays, anything from emails to news reports to social media posts could be the product of AI without you realizing it. If that’s not the definition of social disruption, I don’t know what is.
Be that as it may, conversational AI is here to stay. While it’s inevitable that this technology will bring its own set of challenges, it also invites us to engage in a dialogue with the future. For the first in human history, we can talk to our computers and our computers can talk back at us.
Join the conversation 🗣
Leave comment with your thoughts. Or like this article if it resonated with you.
Get in touch 📥
Have a question? Shoot me an email at jurgen@cdisglobal.com.
I am so excited to get my hands on some different interfaces.
I personally think one of the major obstacles to AI adoption in schools is the chatbot interface feature.
I sense that AI's realtime listening skills lag behind its talking skills. A successful intuitive UI will require AI to understand free flowing conversation in a range of regional accents and dialects. I don't think we're quite at TARS level yet.