Google has introduced software that allows non-programmers to create relatively simple Android applications. The program wraps pre-written pieces of code in bite sized visual representations that can be linked together to create complex behaviors. The software can tap many areas of the Android API, including hardware functions like the accelerometer, and can autonomously respond to stimuli like incoming calls or texts.
The process appears very similar in concept to Apple's Automator, a visual scripting program for Mac OS X. Automator lets users string together a series of actions, essentially creating "to-do lists" for the computer. However, its action library is relatively limited and is best suited to batch operations which a human could do, but wouldn't want to because of time or tedium. Google's application, by contrast, allows scripting to enter the realm of actions humans can't perform - like auto-replying to texts or responding automatically to changing conditions in physical or digital space.
For all its benefits, though, I don't see this program gaining widespread attention immediately. There's still a relatively small overlap between "people who want to build apps" and "people who don't know how to build apps". Outside that intersection, Google's application has little value.
On the other hand, the former category (people who want to build apps) is probably growing every day, and could reach critical mass where the motivation to have X functionality will be answered with Google's easy software solution. From the other direction, as software like this continues to become more advanced and incorporates more complicated scenarios, developers who do know how to build an app may migrate toward the prepack solution. If it produces the same result in less time, why wouldn't they? A majority of smartphone applications are nothing more than distilled tables of a larger database; if Google has an easy way of creating that product then more power to them.
This marks another big step forward in programming literacy. As more of programming's nuances can be wrapped up and handled behind the scenes, it becomes accessible to more people in an imperative form. A person can literally tell his computer/smartphone to do X, Y, and Z, and it will -- provided that the actions conform to the library of easy commands that Google has exposed. Outside those wrappers, developers are on their own -- but just as the visual GUI replaced the command line, simplified programming will become a critical skill as we require our computers to do more than just store files and load the internet.
I remember the first time I "asked" a computer to do something and the pleasant surprise when it returned an answer (no, it wasn't "hello world"). Hopefully applications like Google's will allow that feeling to become democratized and let a larger number of people create applications and extend their computers as active tools. There's no reason a processor has to be stuck loading the internet and connecting calls; it can do just about anything we can imagine, as long as we have some way to express that desire.