I'd just note that blind evaluation could be hard because of systematic discrimination biasing how you evaluate a candidate. For example, if you look at who ha

I'd just note that blind evaluation could be hard because of systematic discrimination biasing how you evaluate a candidate. For example, if you look at who has spoken at conferences before in your blind evaluation, it's blind, but biased by previous discrimination. it's the same sort of issue that crops up regularly in machine learning, where people assume that the data isn't the result of bias. I'd say the correct way to avoid bias is to set a blind baseline of expectations (Like 3+ years in industry working with the topic) and then reach out to minorities. That said, even someone who doesn't have experience can be a good presenter if given the right support, I've seen great conference talk made by total newbies to a technology who were encouraged to write code making use of a new feature. They presented what they learned, problem spots that they found that might trip up someone new to the technology, and open sourced their app as a learning tool.