My most important takeaway from this week's class is that the focus of traditional accessibility technologies is for getting jobs done, rather than for other purposes, including entertainment. Trying to dig into why this is, I think that it's because the incentive alignment is different than that of "normal" technologies.

Two analogs come to mind. First, a common refrain about why enterprise/school software is harder to build than B2C software -- is that the buyer isn't the user. Enterprise software often comes with a lot of features that no one will actually use but that are table stakes before any purchasing manager will approve the purchase.

Second, there's some critical analysis of why SNAP (the program that gives money for food to low-income people) works better than food banks (which directly give food to low-income people) -- it's that food banks create a redundant system of food storage. (An NPR article that gives more context.)

We already have a system called currency that allows people to incrementally convey wants and needs to people who can make things. It's not perfect, but when people try to sidestep the process by having a middleman convey desires to the people who are in charge of providing or building products, a lot of useful information gets inadvertently filtered out. For B2B software, what the purchasing managers have filtered out are concerns like "this feature that looks useful isn't built well and will never be used". For food banks, what policy creators have filtered out is that whatever is lost to abuse of SNAP is so small compared to what SNAP saves by leveraging our existing economic system.

And for accessibility technologies, the message that is lost is that users of accessible technologies are human too. And what every human needs is not just to work, but also to have a good time. Maybe taking a look at the existing analogs can help us come up with a better incentive loop for creating accessibility technologies.