“Let’s train humans first….before we train machines”
- Opinión
We humans are at the absurd stage in our technological evolution when we seem to have abandoned our common sense. Billions are spent by governments, corporations and investors in training computer-based algorithms (i.e. computer programs) in today’s mindless rush to create so-called “artificial” intelligence, widely advertised as AI. Meanwhile, training our children and their brains (already superior to computer algorithms) is under-funded, schools are dilapidated, sited in run-down, often polluted areas while our teachers are poorly paid and need greater respect. How did our national priorities get so skewed?
In reality, there is nothing artificial about these algorithms or their intelligence, and the term “AI” is a mystification! The term that describes the reality is “Human-Trained Machine Learning”, in today’s mad scramble to train these algorithms to mimic human intelligence and brain functioning. In the techie magazine WIRED, October 2018, we meet a pioneering computer scientist, Fei-Fei LI, testifying at a Congressional hearing, who underlines this truth. She said, “Humans train these algorithms” and she talked about the horrendous mistakes these machines make in mis-identifying people, using the term “bias in—bias out” updating the old computer saying, “garbage in—garbage out”.
Professor LI described how we are ceding our authority to these algorithms to judge who gets hired, who goes to jail, who gets a loan, a mortgage or good insurance rates — and how these machines code our behavior, change our rules and our lives. She is now back at Stanford University after a time as an ethicist at Google and has started a foundation to promote the truth about AI, since she feels responsible for her role in inventing some of these algorithms herself. As a celebrated pioneer of this field, Professor LI says “There’s nothing artificial about AI. It’s inspired by people, it’s created by people and more importantly, it impacts people”.
So how did Silicon Valley invade our culture and worldwide technology programs with its short-term, money -obsessed values: “move fast and break things”; disrupt the current systems while rushing to scale and cash out with an IPO? These values are discussed by two insiders in shocking detail, by Antonio G. Martinez in “Chaos Monkeys” (2016) and Bloomberg’s Emily Chang in “Brotopia” (2018). These authors explain a lot about how training these algorithms went so wrong: subconsciously mimicking their mostly male, misogynist, often white entrepreneurs and techies with their money-making monopolistic biases and often adolescent, libertarian fantasies.
I also explored all this in my article “The Future of Democracy Challenged in the Digital Age”, CADMUS, October 2018, describing all these issues of the takeover by AI of our economic sectors; from manufacturing, transport, education, retail, media, law, medicine, agriculture, to banking, insurance and finance. While many of these sectors have become more efficient and profitable for the shareholders, my conclusion in “The Idiocy of Things”critiqued the connecting of all appliances in so-called “smart homes” as quite hazardous and an invasion of privacy. I urged humans to take back control from the over-funded, over-invested, over-paid computer and information science sectors too often focused on corporate efficiency and cost-saving goals driven by the profit targets demanded by Wall Street.
I have called for an extension of the English law, settled in the year 1215: “habeas corpus” affirming that humans own their own bodies. This extension would cover ownership of our brains and all our information we generate in an updated “information habeas corpus”. Since May 2018, European law has ratified this with its General Data Protection Regulation (GDPR), which stipulates that individuals using social media platforms, or any other social system do indeed retain ownership of all their personal data.
So, laws are beginning to catch up with the inhuman uses of human beings, with our hard-earned skills being used to train algorithms that then replace us! The computer algorithm trainers then employ out of-work people surviving in the gig economy on Mechanical Turk and Task Rabbit sites, in minimum, hourly- paid data entry tasks to train these algorithms!
Scientist Jaron Lanier in his “Ten Arguments for Deleting Your Social Media Accounts Now” (2018) shows how social media are manipulating us with algorithms to engineer changes in our behavior, by engaging our attention with clickbait and content that arouses our emotions, fears and rage, playing on some of the divisions in our society to keep us on their sites. This helps drive ad sales and their gargantuan profits and rapid global growth. Time to rethink all this, beyond the dire alarms raised by Bill Gates, Elon Musk and the late Stephen Hawking that these algorithms we are teaching will soon take over and may harm or kill us as did HAL in the movie “2001”.
Why indeed are we spending all this money to train machines while short-changing our children, our teachers and schools? Training our children’s brains must take priority! Instead of training machines to hijack our attention and sell our personal data to marketers for profit — let’s steer funds into tripling efforts to train and pay our teachers, upgrade schools and curricula with courses on civic responsibility, justice, community values, freedoms under habeas corpus (women also own their own bodies!) and how ethics and trust are the basis of all market and societies.
Why all the expensive efforts to enhance machine learning to teach algorithms to recognize human faces, guide killer drones, falsify video images and further modify our behavior and capture our eyeballs with click bait, devising and spreading content that angers and outrages — further dividing us and disrupting democracies?
Let’s rein in the Big Brother ambitions of the new techno-oligopolists. As a wise NASA scientist, following Norbert Weiner’s Human Use of Human Beings (1950), reminded us in 1965 about the value of humans: “Man (SIC) is the lowest-cost, 150 pound, nonlinear all-purpose computer system which can be mass-produced by un-skilled labor”, quoted in Foreign Affairs, July-August, 2015, p. 11. Time for common sense!
- Hazel Henderson D.Sc.Hon., FRSA, is founder of Ethical Markets Media, LLC and producer of its TV series. She is a world renowned futurist, evolutionary economist, a worldwide syndicated columnist, consultant on sustainable development, and author of The Axiom and Nautilus award-winning book Ethical Markets: Growing the Green Economy (2006) and eight other books.
http://www.other-news.info/2018/12/lets-train-humans-first-before-we-train-machines/
Del mismo autor
- “Let’s train humans first….before we train machines” 05/12/2018
- Inteligência Artificial para quem? 05/12/2018