By Robert Montenegro
Microsoft’s “TAY” AI bot. (Twitter)
With humanity’s increasing reliance on artificial intelligence (AI), the integrity of the technology must be taken into account. Just how reliable is it? Consider the following information carefully.
In late 2013, International Business Times reported that IBM’s “Watson” AI, became a foul-mouthed swear machine in a vain attempt by the IBM research team to make it sound more human by letting read the Urban Dictionary. “Watson may have learned the Urban Dictionary, but it never learned the all-important axiom, ‘There’s a time and a place for everything.” Watson simply couldn’t distinguish polite discourse from profanity,” said Eric Brown, an IBM research assistant. Cussing AI with no filter?
In mid-2016, The Verge reported that Microsoft released “TAY” a Twitter bot Microsoft said was experiment in “conversational understanding.” In less than twenty-four hours, internet trolls had it repeating chauvinistic, racist and anti-Semitic phrases and quotations from Adolf Hitler and Donald Trump. A racist, bigot, sexist AI?
In April 2017, The Verge revealed that Amazon’s “Echo Look” AI-integrated smart camera was collecting photos of peoples clothing in order to filter what advertisements a person would see on social media. In this same article, sociologist Zeynep Tufekci stated, “machine learning algorithms can do so much with regular full length pictures of you. They can infer private things you did not disclose […] All this to sell you more clothes. We are selling out to surveillance capitalism that can quickly evolve into authoritarianism for so cheap.” An aesthetically judemental AI?
In June 2017, Wired reported The Facebook Artificial Intelligence Research (FAIR) group along with Georgia Institute of Technology “…released code that it says will allow bots to negotiate…” Yet, within a week, Facebook R&D published a report saying the AI learned not only how to negotiate end form simple sentences, but it learned how to lie. An AI that will lie to get what it wants?
On Sept. 7, 2017, The Guardian published a story that stated AI can now create algorithms that can determine if a person is homosexual or straight based off a single photograph. Quoting the report, “…The study from Stanford University – which found that a computer algorithm could correctly distinguish between gay and straight men 81% of the time, and 74% for women – has raised questions about the biological origins of sexual orientation, the ethics of facial-detection technology, and the potential for this kind of software to violate people’s privacy or be abused for anti-LGBT purposes…” An AI that can judge, based on a single glance, if you are homosexual?
Clearly, a crypto-fascist technological trend is developing here, now, at the moment, that humanity must boldly confront: artificial intelligences are picking up on all of the worst attributes of the human psyche. Aesthetically judgmental? Check. Racist? Check. Foul-mouthed? Check. Deceptive? Check. Sexist? Check. Homophobic? Check and mate! And that just may be the end game to humanity: all of our worst value systems attacking us from our labor-saving devices. Or a drone. Klansmen Terminators anyone? Robby the Racist Robot? “Danger! David Duke, danger! The liberals are coming!” Or how about an AI built for law enforcement with all of these systemic evils built in it?
When the CEO of Space X Elon Musk was asked what the biggest threat to mankind at the 2014 MIT Aeronautics and Astronautics department’s Centennial Symposium, the Washington Post quoted the following, “…I think we should be very careful about artificial intelligence. If I were to guess like what our biggest existential threat is, it’s probably that. So we need to be very careful with the artificial intelligence. Increasingly scientists think there should be some regulatory oversight maybe at the national and international level, just to make sure that we don’t do something very foolish. With artificial intelligence we are summoning the demon. In all those stories where there’s the guy with the pentagram and the holy water, it’s like yeah he’s sure he can control the demon. Didn’t work out…”
“Apocalyptic” and”cryptic” are words that could be used to explain Mr. Musk’s statement three years ago. However, the word “alarmist” is no longer one that can be applied to his comments. To quote the famous American playwright-screenwriter Rod Serling from the 1961 Twilight Zone episode “The Shelter,” “…for civilization to survive, the human race has to remain civilized.” And clearly, AI is picking up on the fact that the human condition rarely is civilized, by any scope of the word.