Yesterday, we announced a new tool that makes it much easier to test out Alexa Voice Service with different hardware and ultimately, to make it easier for third party hardware makers to Alexa-enabled their products. We’ve built several tools and the AVS Kit is our latest.
We can probably breakdown technologies into two categories: fundamental and supportive. Fundamental could be the transistor while supportive are the tools to build, encase, or use them. It’s 3D printing vs the 3D printer, or the filament spool… or its holder. Eventually, you can use supportive technologies to bundle together enough fundamental ones to create a new fundamental technology.
The reason that technological change accelerates is because tools make it easier to build new technologies. In voice, the fundamental technologies are things like Hidden Markov models or deep neural nets, while the tools are the STT, TTS, and other API. In our case, the the AVS Kit removes many manual steps to enable AVS, which itself is a collection of a bunch of technologies.
I can see that in two years, even our one line install is going to appear cumbersome. That’s why planning for 2–3 years out of development is becoming more difficult (and a signal of the Singularity).