Deep Learning & Natural Language Processing: Transformers

Perhaps the most sophisticated, powerful, ambitious, full of potential use of these techniques is as an amanuensis, a digital assistant. Here we imagine our content generation platform playing an integral role with our writer, being a co-author, in a sense, and an integral part of the writing process from the start.

While some writing may simply involve the typing in of a few keywords followed by the full automatic generation of an article, such Quakebot or a stock report, we believe there will always be a place for more nuanced and detailed human input in the process of writing and that our algorithms will always have a place in that place.

Thus, we eventually envision our algorithms being part of a plattofmr that forms a fully co-equal partnership between writer an artificial intelligence, where writer writes down some notes, which trigger a response from AI, which trigger a responder from writer, and so on.

AI has the advantage of being able to scan large masses of information and summary them nearly instantaneously, so that a lot of the heavy lifting of research that humans are now forced to do can be done behind the scenes, without the human even having to direct it—e.g., the AI automatically figures out which parts of the document are best severed by filling in the blanks and doing research.

Thus, a synergy is set up between human and AI writer which takes advantages of the strengths of each of the two.