SCIART MAGAZINE
  • Magazine
  • About
    • Team
    • Contribute
    • Contact
    • Advertise
    • SciArt Initiative
  • Subscribe!
  • Magazine
  • About
    • Team
    • Contribute
    • Contact
    • Advertise
    • SciArt Initiative
  • Subscribe!

DECEMBER 2017

back to table of contents

ON VIEW | INTERVIEW

"Mental Work" by Jonathon Keats at EPFL, Switzerland
Picture
"Mental Work" by Jonathon Keats. Photo credit © Adrien Baraka.
MENTAL WORK: AN OVERVIEW
by Jonathon Keats
 
I. The Context
  • With the Industrial Revolution of the 18th and 19th century, machines outperformed the human body, replacing muscle power in the field and factory. To avoid obsolescence, humans became mechanical underlings tasked with supporting the machines' steam-powered productivity.
  • With the Cognitive Revolution of the 21st century, machines are beginning to outperform the human brain, and may supplant mental power in the office and laboratory, which naturally provokes a number of challenging questions: Can humans avoid obsolescence by supporting machines computationally? Must we? Are there alternative relationships between human and machine? Are there alternative outcomes for the Cognitive Revolution?
Picture
"Mental Work" by Jonathon Keats. Photo credit © Adrien Baraka.
II. The Concept
  • Mental Work offers a vision of future human employment by hiring people to support machine operations with human brainpower. The machines are slider cranks, devices from the Industrial Revolution that not only provided mechanical labor (most famously instantiated in Watt's steam engine) but also calculated mathematical functions (as in the case of mechanical linkages used to approximate trajectories in ballistics). Replacements for human muscle, they are also rudimentary forerunners to present and future electronic computers.
  • The facility has three BCI-operated machines operating at three levels of technological advancement. In physical terms, they are increasingly complex mechanically (and therefore mathematically): While Machine I has no gearing at all, Machine II has two gears and Machine III has three gears plus two linkages working in tandem.  The machines are also increasingly complicated in terms of their brain-computer interactions. Over the course of their hour-long employment, mental workers will progress from the most basic to the most advanced.
  • The work performed by the machines is purely mathematical, the mechanical equivalent of cogitation. The mechanisms are completely detached from the factory, decoupled from any industrial activity. However, by the end of employment on these machines, the mental worker will discern another form of work that has been performed: The machines are mental aids, state-of-the-art versions of the philosophical instruments that helped natural philosophers make sense of the world from the Enlightenment through the Victorian Age. Much as orreries helped to advance astronomy (and provoked profound metaphysical questions about the 'clockwork' universe), these mechanical prostheses help their operators to work out what kind of hybrid human-machine future they want to work for. 
Picture
"Mental Work" by Jonathon Keats. Photo credit © Adrien Baraka.
III. The Experience
  • The training booth trains the future mental worker to work on the three futuristic machines. Simultaneously it trains the machinery to operate in response to the mental activity of the individual mental worker. Conceptually it establishes BCI as a relationship between the computer and the human operator. The nature and implications of this relationship will be more fully explored through the progression of machine interactions.
  • Machine I allows the human operator to turn the machine on and to alter the speed with brain activity. Conceptually it establishes a naive baseline for interactions with Machines II and III by assigning cognitive control of a machine to the brain of a human operator via a standard BCI interface.
  • Machine II has the capability of providing a greater level of control, but greater control demands greater effort – possibly greater than the human brain can achieve. The degree of control can be mentally modulated by a human supervisor in response to the operator's performance, and is adjusted by changing the probability threshold (or in other words the strength a brain signal must have for the machine to execute a command). As a result of this relationship between worker and supervisor, the black box of the BCI is opened up. A hidden layer of the BCI algorithm is exposed. Conceptually Machine II challenges the assumptions made by the naive mental worker on Machine I, questioning whether the operator's brain is ever truly in control, and showing the influence of the interface.
  • Machine III allows the supervisor to select the operator's brain-computer interface via BCI, choosing between motor imagery (MI) and Alpha wave activity depending on the operator's state-of-mind as gauged by the performance of the machine. MI has the benefit of giving the operator fully-alert control, while Alpha offers a work regimen within the capacity of a mentally-exhausted human operator. (The onset of mental relaxation that correlates with an increase in Alpha activity is inherently less precise than a specific motor signal, but this disadvantage may be outweighed by the advantage that relaxation is inherently easier on the brain.) Conceptually Machine III provides meta-level control over the machine, as represented by the supervisor's choice between interfaces. It reveals that the interface is not absolute, and suggests that no interface is universally optimal: An interface is only as good as the degree to which it fits its context, but the right match can be magnificently empowering. By extension, the Cognitive Revolution is not necessarily a diminishment of our human abilities; it can in fact be an enhancement because the possibilities for interfacing are limitless.
Picture
"Mental Work" by Jonathon Keats. Photo credit © Adrien Baraka.
IV. Some Implications
  • Humans can not only support machines computationally; we also have the power to determine our supporting role, which is the fundamental source of human control. The human-machine power dynamic is a matter of perspective.
  • Creating and selecting the algorithms of interaction is the invisible mental work underlying the Cognitive Revolution.
  • No algorithm is inherently good or bad. Different algorithms may serve different people, and the diversity can reveal a diversity of human cognitive abilities: Depending on the interface, someone with high body awareness may be more mentally adept than an intellectual. The diversity revealed through the Cognitive Revolution may therefore be humanizing.
  • The relationship between the algorithm and the user is fluid. While our brains have the power to change the interface, the interface has the power to change our brains (on account of brain plasticity and ultimately genetics). Underlying the Cognitive Revolution is human-machine co-evolution.
  • The mental work of deciding how machines interact with humans is a fundamental human responsibility. The responsibility was essentially abdicated during the Industrial Revolution with the advent of machines such as Watt's steam engine. Today, as we face a future of BCI, we ignore the foundations of human-machine interaction at our own peril.
Picture
"Mental Work" by Jonathon Keats. Photo credit © Adrien Baraka.
Julia Buntaine, SciArt Magazine: Your exhibit proposes, in essence, to abandon the body and use the power of the mind as a means of remaining useful as machines automate away our jobs. This naturally leads to the thought of the cyborg, or biological mergence with technology. If the opportunity arose, what would be your dream cyborg-enabled attribute?
 
Jonathon Keats: It's easy to think of ways in which a cyborg future might make life easier. For instance, I'm a very slow typist, and I could easily be seduced by a brain-computer interface that would automatically transcribe my thoughts. But I am wary of this way of thinking about cyborg enablement, because it's also disabling in the sense that my machine-enabled self would merely diminish whatever was human in me. One way around this would be to completely abandon the distinction between my biology and technology. In that case, I might imagine how my cyborg self might be more creative for the fact that menial tasks like typing would be made more efficient, no longer requiring the roundabout transfer of my thoughts through finger muscles and the spring-activated electrical contacts on my keyboard. But there are a lot of questionable assumptions in that proposition. (For instance, might creativity actually emerge from inefficiency?)
 
So while I am open to experimenting with technologies – and am convinced that we've been cyborgs since the invention of the handaxe – I'm actually more interested in considering how a cyborg future might make life more challenging, enhancing my humanity. What I mean by this is that our technologies have already vastly augmented our abilities. Our technological capabilities have progressed at a rate well beyond human evolution, and we're simply not evolved cognitively to use the powers we have at hand with adequate responsibility. I believe this is one of the principal reasons why the world we've built is so precarious. Therefore I'd like to explore technological prostheses that make us reexamine ourselves, and by those means make us less certain, less decisive, less efficient. In a sense that is my motivation for enlisting brain-machine interfaces to operate philosophical instruments in the Mental Work facility. This is a step in the direction of a cyborg future, but these are machines to make you stop and think. 
Picture
"Mental Work" by Jonathon Keats. Photo credit © Adrien Baraka.
JB: In the science fiction future where money may be done away with altogether (as seen in the positivist vision of Star Trek), it seems to me that Mental Work could have a nice partnership with your Spacetime Industries - paying workers in ingots of time. For now, however, your viewers will get the satisfaction of "being ahead of the curve" in the creation of their Mental Work. Do you have plans to collect or record the work that is made?
 
JK: Mental Workers are not paid in cash. Their compensation for their work is cognitive. The machines induce thought, and workers' thoughts belong to the workers to do with as they choose. Workers will not be required to reveal anything. However, in my opinion, thoughts increase in value when they're shared. (In that sense, they're manifestly unlike cash, even if this difference tends to get overlooked by intellectual property law.) So sharing will certainly be encouraged, whether through voluntary surveys or spontaneous conversations (which may be first- or second- or third-hand). Ultimately my ambition with the Mental Work facility is that it becomes a hub for contemplating, discussing and debating possible approaches to the cognitive revolution, enriched by the experiences of all participants and their encounters with people who may never have the chance physically to be connected to our machines.
 
In tandem with the mental work undertaken by machine operators, scientific data will be collected by José Millán, my scientific partner. His lab will systematically analyze human interactions with the machines, and the raw data – which will be completely anonymous – will eventually be made available to researchers worldwide. Given the sheer number of workers expected to pass through the factory, and their diversity, the data set will be unprecedented in the field of brain-computer interfaces. The research will open up new possibilities for the technology at the same time that public debate about the technology will increase.  
Picture
"Mental Work" by Jonathon Keats. Photo credit © Adrien Baraka.
JB: I personally find the idea of artificial intelligence making art fascinating, and am intrigued by how your exhibit subverts this notion. So, can we paint in the Mental Work facility?
 
JK: While the Mental Work facility is not outfitted for painting, there is no reason why one could not paint with a brain-computer interface. A robotic arm could hold the brush, and commends could be sent via an EEG headset. That said, BCI control is very rudimentary at this stage. It would take an extraordinary amount of effort to paint a nude or a bowl or fruit or even a simple black square. The process would be highly inefficient. However, as I suggested above, creativity might well emerge from inefficiency. The struggle of painting with brain signals might lead to new aesthetic discoveries, much as the challenge of operating machinery in the Mental Work facility may generate novel thoughts.
Picture
"Mental Work" by Jonathon Keats. Photo credit © Adrien Baraka.

SciArt Lifetime Digital ​Subscription

$50.00
Buy Now

Subscribe once, be set for life! One time payment, no renewals.

​Upon purchase, your digital access code will be automatically emailed to you.

For gift purchases, simply forward or print out your confirmation email.

SciArt Magazine is a publication of
​SciArt Initiative, Inc.