Alter has a genderless, ageless face. Its arms resemble those of a human being, but it has the bare body of a machine. Yet it is learning and building memories that will assemble into a “personality.”
Or so its co-creator Takashi Ikegami seems to think. "The mind is not generated from within, it is always coming from outside," Ikegami, Artificial Intelligence (AI) researcher at Tokyo University, told The National. "So the question is how to copy the mind?"
The interaction-based android is one of the cutting-edge research projects presented at a new exhibition at London’s Barbican that reflects on the ways AI will impact our very existence.
AI: More than Human sets out to challenge "preconceptions of AI" by engaging visitors head-on with the latest innovations. Whether these preconceptions are an inordinate excitement for all things AI or a weariness of them, the exhibition offers a chock-full of new stimuli.
Alter is perhaps the most mind-boggling creation.Ikegami speaks of the machine as if it shared human traits, citing a "warm human-android interaction" and calling "memories" the information gathered through the internal simulator that sits in its humanoid eyes.
By means of this "learning" process, Ikegami thinks the mind can be "offloaded from a human being into a robot" – not too dissimilarly from how a toddler learns to be part of society though observation.
Alter has already successfully conducted a philharmonic orchestra in the first "android opera" – fittingly titled Scary Beauty.
But while the story of Ikegami eerily resembles the literary one of Victor Frankenstein, a young scientist who creates a hideous, sapient creature in an unorthodox scientific experiment, he is undeniably driven by a deep love for science and the conviction that "human and android can work together to understand something which human beings have not yet been able to understand."
The exhibition does not shy away from the hard questions conjured by dreams of this magnitude: Will what we find be beneficial? And will ethics be sacrificed in the process?
When AI goes wrong
Tay, a Microsoft creation, is one example of where AI has gone wrong. The chatter bot was taken offline after it was taught within days by users on Twitter to say racist things, but at the Barbican it is still spewing its racist slur.
Deepfakes of former US President Barack Obama also demonstrate how easy it is to trick a viewer into believing misleading footage manipulated through artificial intelligence. The Campaign To Stop Killer Robots even has a small section of its own, featuring explanatory material on why it is prudent to pre-empt autonomous weaponry on the battlefield – at least until we find a good way of putting human values into machines.
The interest in AI peaked in 2010, when machines increased by leaps and bounds thanks to fast processors that allow the analysis of huge amounts of data. But the idea of the artificial creation of life has ancient roots.
The exhibition recalls ways in which humanity has always been intrigued by the artificial creation of living beings, whether through magic, science, religion or illusion. The desire to communicate with inanimate objects has expressed itself differently across civilisations – from the religious traditions of Shintoism and Judaism to the science of Arabic alchemy – but was always driven by the attempt by humans to explore their place in the world.
How AI could save the world
Humanity may now be looking at AI not only to explore its place in the world, but to fix it.
MIT Media Lab’s Open Agriculture Initiative prompts considerations on what the “next agricultural revolution” will look like in light of climate change.
John Oswald, Managing Director at Method (which is affiliated to MIT) thinks a tabletop-sized, controlled-environment technology farm will be the future of agriculture.
The doom and gloom scenarios that he and his colleague David Mayman describe include the depletion of all nutrients from the earth’s soil by 2050, making traditional farming virtually impossible.
They believe the solution to the problem is a high-tech transparent portable cube that everyone will be able to grow vegetables in. "I call it edge farming: you are taking it closer to where consumption actually happens," Oswald told The National.
Mayman called it instead an “acceleration of hydrophonics” and stressed that the box components are recycled to enhance sustainability (despite it being hard to imagine what environment will be left to protect).
Testing it out
Sony also presented its recipe for the future with its Kreyon City, an interactive installation where users are challenged to create the perfect city. The goal is both an educational one for the audience and a scientific one for its creators.
It does not take long for the user to realize that adding more green lego bricks (for parks) or red ones (for schools) does not necessarily bump the "quality of life" tally on the screen. "For every attempt, we are recording information on how the person is learning the concept, whether there are common strategies and if the interaction with the machine evolves as they play," Enrico Ubaldi, assistant researcher Sony CSL, told The National.
One day, perhaps, the algorithm Ubaldi and his colleagues crafted will come up with a clever way of changing how we run our economy that fixes climate.
Until then, visitors are left with questions, albeit perhaps better-informed ones.
"I think you would struggle to come out [of the exhibition] without thinking this is really exciting, or this is really interesting," Charlie Beckett, director of the Media Policy Project at LSE, told The National after the visit.
Beckett, whose work currently focuses on the ways in which AI is changing the journalism industry, said that “where there are problems we can try and address them.”
Come what may, the Barbican's exhibition may be to date the biggest and most ambitious attempt to start a conversation on how to do just that.
AI: More than Human is at London's The Barbican until August 26
