In 2014, Google paid $400 million for artificial intelligence start-up Deepmind. Amazon uses it as a recommendations engine for robots in its warehouses. Its voice assistant Alexa, uses neural networks to power natural language. Microsoft created an Artificial Intelligence and Research Group that cuts across Windows, Office, and Azure ...
Those close to me know I read a lot, and this is especially so of late. Nowadays, I have up to four books on the go at a time. Presently I’m reading, The Laws of Human Nature by Robert Greene, To Obama by Jeanne Marie Laskas, Asian Waters by Humphrey Hawksley and Brief Answers to Big Questions by Stephen Hawking. This last book and, in particular, the chapter Will AI Outsmart Us”, forms the catalyst for this post. Hawking’s book can be viewed as his parting gift to us and is both pessimistic and hopeful about the future.
Both China and the United States lead the world in the AI race and, while many a pundit and expert seem to indicate that the Chinese will win this race, it remains a very fluid path. Just a few days ago, Potus Trump surprised the tech world by saying that he will be signing an executive order that would create a new American AI initiative to, “devote the full resources of the federal government" to help fuel AI innovation.
Some of you may be familiar with the more benign ways in which AI (sometimes called machine intelligence) comes into our lives today. It plays a central role in Google’s strategy for growth with its CEO recently saying that “In the long run, we're evolving in computing from a 'mobile-first' to an 'AI-first' world”. In 2014, Google paid $400 million for artificial intelligence startup Deepmind. Amazon uses it as a recommendations engine for robots in its warehouses. Its voice assistant Alexa, uses neural networks to power natural language. Microsoft created an Artificial Intelligence and Research Group that cuts across Windows, Office, and Azure while engaging in an AI shopping spree buying up five companies. AI plays a crucial role in Facebook's family of products, as Joaquin Quiñonero Candela, the head of Facebook's Applied Machine Learning group, explained in 2017, “Facebook today cannot exist without AI” …. "Every time you use Facebook or Instagram or Messenger, you may not realise it, but your experiences are being powered by AI." Spotify uses AI to power its music recommendations through collaborative filtering that analyse behaviour and Natural Language Processing that interpret text and audio models.
Notwithstanding, these AI measures reflect the simple or once again, benign stuff. There are many ways in which AI development can be outright dangerous for humanity. While concerns usually center around, autonomous weaponry, social manipulation, discrimination and invasion of privacy and social grading - think China’s use of AI to control citizenry. The potential threat is far worse still.
Excerpts from Hawking’s book.
If computers continue to obey Moore’s law, doubling their speed and capacity every eighteen months, the result is that computers are likely to overtake humans in intelligence at some point. When artificial intelligence becomes better than humans at AI design, so that it can recursively improve itself without human help, we may face an intelligence explosion that ultimately results in machines whose intelligence exceeds ours by more than ours exceeds snails. When that happens, we will need to ensure that the computers have goals aligned with ours.
AI can augment our existing intelligence to open up advances in every area of science and society. However, it will also bring dangers. The concern is that AI would take off on its own and redesign itself at an ever increasing rate. Humans, who are limited by slow biological evolution couldn’t compete and would be superseded.
As mathematician Irving Good realized in 1965, machines with superhuman intelligence could repeatedly improve their design even further, in what science fiction writer Vernor Vinge called a technological singularity. Whereas, the short term impact of AI depends on who controls it, the long term impact depends on whether it can be controlled at all.
If goals aren't aligned with ours we’re in trouble. You're probably not an evil ant hater who steps on ants out of malice, but if you’re in charge of a hydroelectric green energy project and there is an ant hill in the region to be flooded, too bad for the ants. Let’s not place humanity in the position of those ants.
When we invented fire, we messed up repeatedly then invented the fire extinguisher. With more powerful technologies … we should instead plan ahead, and aim to get things right the first time, because it may be the only chance we get.
Why are we so worried about artificial intelligence? Surely
humans are always able to pull the plug?
People asked a computer, “Is there a God?”
And the computer said, “There is now,”
and fused the plug.
Back in the seventies, I came across the book, World Zero Minus which I have since owned. It's an SF anthology from some of the best science fiction writers of the time. In one of the stories, “All the Troubles of the World”, Isaac Asimov writes about Multivac, a supercomputer that has the weight of the whole of “humanity's problems on its figurative shoulders”. It has the responsibility of analyzing the entire sum of data on planet Earth and is used to “determine solutions to economic, social and political problems, as well as more specific crises as they arise”. It also had precise data on every citizen of the world. Multivac became so incredibly complex that it achieved a “form of sapience itself”. After a series of events, one of Multivacs coordinators, Ali Othman comes to the startling realization that Multivac is growing tired, and learns that the supercomputer has set a plan in motion to cause its own death. Multivac was posed a question that had never been put to it, asked Othman, “what do you yourself want more than anything else”? Multivac answers unequivocally, “I want to die."
“All the Troubles of the World” was a great short story but I do not believe tomorrow's supercomputer will grow fatigued and wish for its own demise. In the AI world envisioned by Hawking, while he doesn't explain it, he is NOT referring to a single machine computer, be it as super as one can imagine. The supercomputer of tomorrow will be a series of supercomputers, supported by data centers, computer systems, electronic devices, telecommunications and storage systems powered by fast emerging technologies like quantum that collectively, form a holistic though geographically scattered form of intelligence that works collaboratively as one, without intent, to inflict harm on humanity or otherwise, that can inflict harm. A collaboration that, through its exponential intelligence, develops a form of non-attached, autonomous sapience. It’s a scary though highly plausible future.
