Ok, first addressing the common complaints:<p>- Yeah, DAWs suck, poor example, let's ignore this.<p>- The title is probably not very well chosen, as some other commenters highlighted by focusing a lot on creativity on itself. The actual piece seems to try to focus more on making computers better platforms for the practice of creativity, not so much about the creative limits of computers on themselves.<p>- This is a technical crowd, and the piece is more about "what would be nice", so no point bothering that much on technical issues.<p>I actually agree with the main point, but indeed, just saying "what would be nice" is not that helpful. For the point on standarization, I think a better, more technical and actionable framing, would be to try to port the concept of <i>type safety</i> from languages to specs and APIs. While there's some kind of informal consensus on the main types that are used everywhere, if we could (yeah, this is a massive leap, but I'm only trying to illustrate the idea) pass "specs" into functions in type-safe languages, just like we can pass complex functions as parameters in many modern languages, and we could automatize to some degree compliance and testing, that would be awesome. So, not so much about standarization, but about a new abstraction level, making the interoperability of computational systems type safe. I think about something as common as urls, and I tell to myself: this is such an unsafe mess. Even if a spec is technically well defined, translating that into code is too much.<p>The other general point, moldability, might be better expressed as "accessibility". Making systems more accessible to users. But I think besides many poor software tools made in a rush for specific interests, we are already doing decently here when we really try. It's only that big companies that make the big products usually have too much inertia to continue doing certain things the same way they have been done in the last 40 years (most DAWs are a good example as many commenters pointed out), and smaller actors don't have enough resources to make complex tools that are much cooler than what we have. But on this aspect, I don't think the obstacles are that significant.<p>Oh, and on a final note, nice writing (even if it wanders a bit too much sometimes) and really beautiful design. Even the html is pretty readable.