We’re now over a year in and so far it seems like an incremental improvement. It’s useful for doing one-off things that are outside of normal development, but for most of my daily work it’s close to useless. I certainly haven’t seen anything like 10x productivity.

Those that adapt to the new tools may more than 10x their productivity, in which case this market would resolve as NO, even if it's not true for all software engineers

Very interesting prediction actually.

What if they're mostly or completely automated due to AI? I'd vote that the spirit of this market would say that this scenario would resolve to NO, as the thing being tracked is AI influence on the workforce, and it would be weird to have the resolution "wrap around" for too much AI influence

@Nikola

Yeah, in that case would resolve to NO

Does it need to be obvious that there are no software engineers that are 10x more productive, or is the resolution YES if it's obvious that most software engineers aren't 10x more productive?

My arguments in favor of NO:

• In some cases I'm already being twice or three times as productive than I was in 2022. Some of it is due to copilot and ChatGPT, and some of it is due to other tooling improvements.

• The resolution criteria doesn't specify that the 10x improvement needs to be solely due to AI. There are amazing productivity improvements in programming languages, libraries, services, cloud computing, and developer tools that are unrelated to AI.

• The way I understand it, if there is ambiguity around whether or not developers are 10x more productive, the resolution is NO.

@YoavTzfati Note - I'm a professional software engineer.

@YoavTzfati The problem with that argument to me is the bottlenecks. Even if you make a task infinitely more efficient, if that task was only 5% of your job then you’re only 1.05X more efficient. If more than 10% of your work is meetings then becoming 10X more efficient requires the efficiency to also help your meetings.

With the other developer tooling, it doesn’t seem to me like software development is even just 1.5X as efficient as it was five years ago. 10X more efficient is an absolutely huge jump.

Also, not to disbelieve you, but are you really 2-3X more efficient than a year ago? You’re completing three times the number of tickets/solving three times as many bugs/etc.?

@Gabrielle I admit that 2x - 3x is on a specific set of tasks and not all of my work. Specifically - I can write scripts and exploratory code in python 2-3 times faster than I did before, and in some cases UI code.

I can definitely believe that AI will help with all pieces of a developer's workflow, including meetings.

Would the author care to share some qualifications to show how in touch they are with the field of software engineering? I'm not sure the layman would be able to adequately judge [the obviousness of] the productivity delta.

@Gigacasting Ooh, a book I've read!

How does this resolve if most software engineers have been replaced by AI

@vluzko in that case I assume software engineer are being more than 10x productive, so resolves NO.

@FranklinBaldo why would you assume that?

@Odoacre because productivity can be measured ratio between output/input, if our case output would be lines of code and the input would be human labor.

@FranklinBaldo that's a stupid measure for many reasons.

@Adam ok, share your wisdom

@FranklinBaldo i selling my shares now. And to avoid conflict of interest with market resolution criteria i will no longer trade in it

@FranklinBaldo forgot about this commitment . Already sold again.

@FranklinBaldo I assume the original version holds?

"in that case I assume software engineer are being more than 10x productive, so resolves NO."

@YoavTzfati Yes