People are talking a lot about vibe coding, these days. What will this mean for people working as software engineers? My own personal sense is that productivity will be greatly improved with LLM tools - something we’re starting to see happen - but it’s premature to say large numbers of jobs will disappear. We'll see how things play out.
An important factor around this trend is “getting the right vibe”. How can we be expected to vibe code when the vibe is off?
We’ve been thinking about this a lot at Autodesk Research, where we’re looking at developing tools to support the vibe coding community. We want to make sure that coders of the future live and work in environments that support their wellbeing and maximize their productivity (there will always be someone else with a better LLM waiting to take their job, after all).
One important aspect of this work is to determine what a room’s vibe consists of, from static properties such as a its materials to dynamic properties such as lighting levels and the time of day. All of these affect a room’s vibe, and therefore your ability to vibe code.
Last summer our research team in Toronto was joined by Tongyu Zhou, a PhD student from Brown University, who created a tool called Vibescaper that allows you to assess and modify the vibe of a space and even perform “vibe transfer” from (for instance) mood images to virtual spaces. Which could in turn be transferred to physical spaces, too, of course.
I predict that much of our vibe coding will eventually take place in virtual environments (and might already be, depending on what you believe), so having a tool such as Vibescaper available is likely to have a strong, positive impact on our ability to develop software.
I don’t usually finish my April 1st posts by revealing their falseness, but I’m a little worried that you might read this far and think that Tongyu’s work doesn’t exist. It does - and expect to hear more about it in far more serious settings - but it isn’t being focused on the needs of the vibe coding community, if there even is such a thing. Thank you, Tongyu, for your hard work with us!
