“If a man takes a run in a forest, and there is no Fitbit to record it, did it happen?” [Matthew De Abaitua 2019]
Matthew De Abaitua is a writer and academic who teaches creative writing at the University of Essex. I caught up with him at the AI and Simulation of Behaviour convention (AISB 2019) in Falmouth, Cornwall. He was honest and disarming – he started his keynote by explaining that novellists are usually disappointing in person, where they are unable to present their redrafted best selves :).
3 of Matthew’s novels explore topics in AI and its futures:
- The Red Men imagines a world where human consciousness can be uploaded to a computer (actually as part of an AI system, Cantor) with a business model that charges for simulation of the outcome of policies and decisions applied to “Red Men”. [Update – I am well into “The Red men” now and will present a spoiler-free blog about it soon].
- IF…THEN considers a future where a Sussex town which generates no economic value is bought by and becomes the plaything of an AI with a penchant for war simulation.
- The Destructives
I was lucky enough to have a chat with Matthew at the conference, where he discussed some of the methods of his craft over a glass of wine: The initial search for a “hook” followed by a huge amount of research to get to know just enough so as to have interesting things to say, but not so much that the writer ends up repeating the prevailing knowledge. Huge numbers of rewrites on a section-by-section basis until it “feels right”, during intensive daily writing sessions (9-3 with a break) followed by complete breaks from the (explicit) creative process. This tallies well with others I have spoken too, as well as the way that I have naturally fallen into during my sabbatical.
Matthew also gave a keynote at the AISB conference which considered a range of aspects of the utility of imagination in AI research and development. He started by underlining the power of the novel as a tool to evoke empathy and new understanding – the reader’s point of view creates the novel on the scaffold of the writer’s words, pointing out the utility of such ideas via a 2017 paper from Google Deepmind which explored imagination-enhanced machine learning.
Matthew was critical of the sloppy use of language in the computational creativity research community – an unfeeling machine cannot create poetry (only verse) so long as poetry is the expression of feeling via diction. In a later question on the creativity involved in the creation of AI algorithms he stated the interesting idea that the major difference between art and technology is that art is allowed to fail but technology is not. My own reflection is that technological failures tend to be on a larger scale than artistic failures – for example the stealing of attention by AI-fuelled social media. Simply that technological failure often results from not being able to see the wood for the trees (while wearing a FitBit while running through the forest :).
He referred to a good body of work by thinkers such as the philosopher Alva Noe’s contention that it is in our nature to reflect on and acquire the nature of other. This provides a compelling argument for using “he/she” in the novel to permit multiple points of view and to allow the exploration of characters through understanding their (possibly mistaken) beliefs about other characters, whereas the use of “I” can lock the author into a single mind.
While English students learn about history through the representation in the novel, it would be fascinating to understand what history looks like through the lens of AI. For example we might argue that early portrayals of robots reflect a failure of empathy. He discussed the portrayal of time in a world of AI and data in novels (in my view the marvellous Einstein’s Dreams is a fascinating fictional exploration of time – which I found through the grace of the good taste of my San Francisco AirBnB landlord :).
On questioning, he ventured that the dominant narrative at the moment is of a chaotic world – which may behoove novelists to show a utopian view, though he feels a dystopian view may more accurately represent futures where advanced technology is plugged into political and economic power.
Hopeful, utopian ventures such as the Augmented Intelligence Summit may go some way to helping achieve that utopia! Certainly I will be working towards that…
You can find Matthew De Abaitua on Twitter @MDeAbaitua