Tag Archives: ai

Thought #58. Checkmating AI.

My dog Sucre is not a dab hand at chess, but he has the uncanny knack of soothing me when necessary. I wonder whether AI (Artificial Intelligence) would ever possess such a humanlike capacity. Last week I came across a series of articles on the future of AI. Among them, it was the one on Erica, a Japanese android created 23 years ago, which riveted my attention.

Back in 1996, the classic man vs. machine plot hit the headlines when Deep Blue, an IBM computer specialised in playing chess beat the world chess champion Gary Kasparov. I clearly remember how exciting it was to see AI in action. I read with rapt attention all the articles that flooded the press in those days.

However, I was disappointed about the unfair comments made by the majority of journalists. Instead of a breath of fresh air, Deep Blue was the spark for a spiral of overthinking that bordered on the absurd. A mounting ill-founded concern over the risks of AI that has remained alive up until now. But neither was Deep Blue, at that time, a menace to humankind, nor does Erica represent a hazard nowadays.

Erica has a beautiful face and can express emotions. She can also be engaged in a conversation on topics ranging from culture, politics or new thinking, to the hypothetical scenario in which AI becomes the dominant form of intelligence on Earth. She has a relatively high degree of autonomy and has been programmed to learn from each human interaction in which she is involved.

When asked about the AI takeover, she openly recognises that despite being difficult to predict, it is too soon to envisage so fascinating a scenario. Many of us, at this moment, would ask ourselves whether doomed would not have been a more appropriate adjective to qualify that scenario. In any case, Erica’s arguments are sound.

She provides a snapshot of the possible risks of AI. In the first place, she assures us that an AI would not be capable of stabbing human beings in the back, unless it was programmed to do so. She also claims that AI’s power of computing is not tantamount to the amazing ability to feel and react that humans have. Somewhat, not a mere sign of a conceited attitude can be appreciated when conversing with Erica.

Of course, she is proud of her creators. She talks with frank enthusiasm about the new algorithms that they are planning to implement in the near future. Apparently, these blocks of code are aimed at enabling Erica to understand and emulate feelings. However, it is rather comical when Erika tries to imitate an awkward silence; she always fails in her attempts to choose the right moment to stop talking. More embarrassing though is when she tries to be funny and tells a cruel joke about genocide.

All things considered, I would decidedly go for a dog. Sucre’s companionable presence can soothe me and even make me happy. I would not go so far as to say that the lack of sense of humour could tarnish Erica’s reputation, but she definitely lacks what makes us human.

Thought #13. AI bias.

Once upon a time, people knew the difference between rightness and badness. As if it was an Opera, the history of AI (acronym of Artificial Intelligence) began promisingly, but then things started to go wrong.

The opening of AI goes back to ancestry when the Greek mythology created Pandora, the first “artificial” woman. Each Olympus god was commissioned to make a contribution to Pandora’s process of creation with the sole aim of taking beauty into account.

During the interlude, step by step, new features and unique gifts were conscientiously added by the band of male creators. They went to great lengths to make Pandora in their own image (a mental image of a sexually arousing creature).

Pandora received a jar as a wedding present and was told not to open it. Thinking that it was a mere misdemeanour, she could not help opening the jar, triggering unexpected consequences.

To the astonishment of the ancient Greek gods (and goddesses) of the epoch, the sad coda was that Pandora ended up in a poetry epoch column (probably from a tabloid) depicted as an archetype and role model of future women generations:

“From her is the race of women and female kind: of her is the deadly race and tribe of women who live amongst mortal men to their great trouble, no helpmates in hateful poverty, but only in wealth.” ― Hesiod, Theogony

Fear for inventions and machinery has been thenceforth present. The Industrial Revolution seemed too profound a change to overcome, but eventually all the criticism levelled at technology advances started to wear off.

AI is a force to be reckoned with but I wouldn’t go so far as to say that it will eventually surpass human capacities. Think about a relatively recent groundbreaking invention: the digital camera sensor. If George Orwell had known about such an invention, Nineteen Eighty-Four would probably never have been written.

Those who try to undermine science (or engineering) by spreading false alarms concerning the so-called perils of AI should be treated with the contempt that they deserve. They seldom use nothing but devious tactics to succeed in turning utter nonsense into a burning issue. Should nonsense be regarded as tantamount to blatant ignorance? Apparently, according to Prof Stephen Hawking and others of his ilk, it should not.

“The development of full artificial intelligence could spell the end of the human race.” ―   Stephen Hawking

“Artificial intelligence is our biggest existential threat.” ―  Elon Musk

Nevertheless, as far as some basic safety rules are not wilfully ignored, AI is harmless, secure and risk-free. Just as women are the most perfectly created sculpture ever chiseled, driverless cars (aka autonomous car) are also something that will become ubiquitous in the forthcoming years.

Do not blame me! My complete disregard for the underminers is based on their own arguments, which should fly in the face of common sense. The inability to comprehend how technology works still rankled with them.

Not being in the mood to reproduce in my own blog more quotes from them, I rather like the idea of presenting you with a biased view of the issue, based on as yet unproven facts:

“Real stupidity beats artificial intelligence every time.” ― Terry Pratchett, Hogfather

“The question of whether a computer can think is no more interesting than the question of whether a submarine can swim.” ― Edsger W. Dijkstra

The rest is history (to be written).