What kind of universe is it that we inhabit, which celebrates itself as a society of choice but in which the only option available to the enforced democratic consensus is a blind acting out?
The true courage is not to imagine an alternative, but to accept the consequences of the fact that there is no clearly discernible alternative: the dream of an alternative is a sign of theoretical cowardice, functioning as a fetish that prevents us from thinking through to the end the deadlock of our predicament. In short, the true courage is to admit that the light at the end of the tunnel is probably the headlight of another train approaching us from the opposite direction.
Designers need a new way of grasping ML’s capabilities, a kind of abstraction that focuses on the match of contextual capability and user value; a kind of taxonomy that is likely to be radically different from ones used by data scientists.
Particularly, three clusters of ML technical advances have not yet been bound to particular utilities, interactions or user experiences. These design-wise underutilized clusters are: deep learning, sentiment analysis and social network mining.
It seems that ML challenges the general idea of prototyping; of making just enough of a system to assess if this is the right direction to go. ML seems to require a much higher level of commitment, requiring an unwieldy amount of data to create a functional prototype. This could conflict with UX mantras like “fail fast, fail often.”
Designers often have cliched understandings of this medium, driven by the hype or criticism surrounding the field. For example, some typical stereotypes are that wearables are for fitness, artificial intelligence is for automating