Fears of #AI Part 2 – Accidental Intelligence

Just as we were recovering from the spat between Mark Zuckerberg and Elon Musk, as discussed in last week’s post on Fears of AI, new events in the world are leading to more fears of perils due to Artificial Intelligence.

Facebook stopped an experiment where it had created two AI Bots to negotiate with each other. The initial idea was for the bots to negotiate with each other, based on bartering. Thus ultimately hitting the common ground for both and close the deal. Instead, this happened –

(Original Transcript of negotiation between the bots)

Bob: i can i i everything else . . . . . . . . . . . . . .

Alice: balls have zero to me to me to me to me to me to me to me to me to

Bob: you i everything else . . . . . . . . . . . . . .

Alice: balls have a ball to me to me to me to me to me to me to me

Bob: i i can i i i everything else . . . . . . . . . . . . . .

Alice: balls have a ball to me to me to me to me to me to me to me

Bob: i . . . . . . . . . . . . . . . . . . .

Alice: balls have zero to me to me to me to me to me to me to me to me to

Bob: you i i i i i everything else . . . . . . . . . . . . . .

Alice: balls have 0 to me to me to me to me to me to me to me to me to

Bob: you i i i everything else . . . . . . . . . . . . . .

Alice: balls have zero to me to me to me to me to me to me to me to me to

What is odd in the discussion above is the language that is being used by the two bots to negotiate. Clearly, it’s not English but something scripted in English. What is clear is that the bots have developed a language only they can understand. They are interacting and responding the same way. This is the classic GIGO (Garbage In Garbage Out) situation. However, Facebook quickly shut the experiment. The whole world has gone berserk with this announcement. Facebook has embraced the greatest fears of AI. Some go to the extent to say this is what has happened in public domain. What has happened in private, no one knows?

There is more noise now that scientists may not know all the solutions, and to be more specific all the problems. There are doomsday scenarios being written by multiple people, but here is a good list that tells all –

1. THERE ARE NO ETHICS, INDUSTRY STANDARDS OR TESTING RULES
2. RELATIONS BETWEEN MAN AND MACHINE IS CHANGING
3. GENETIC DIVERSITY WILL BE WIPED OUT
4. AI IS GOAL-ORIENTED REGARDLESS TO A FAULT
5. WE WILL HAVE NO CONTROL OVER THE MACHINES
6. VALUE OF LABOR AND SKILLS WILL FALL
7. CRIMINAL POTENTIAL FOR ARTIFICIAL INTELLIGENCE IS BIGGER THAN YOU THINK

8. REALITY IS DIFFERENT THAN HOW THE MEDIA REPRESENTS AI

9. WE ARE HANDING OVER THE KEYS

10. IT WILL LEAD HUMAN RACE CLOSE TO THE DEAD END

11. AI CREATORS WILL NEVER KNOW ALL THE SOLUTIONS

12. LAWS AND FUNDAMENTAL RIGHTS WILL LOSE THEIR IMPACT

13. OUR EDUCATION SYSTEM NEEDS TO BETTER PREPARE THE NEXT GENERATION

14. THERE’S NO CURE FOR IGNORANCE

15. AI IS NOT A PROBLEM, BUT HUMAN HUBRIS IS

More details here – The Clever.

Here is the video of the week –