Thread by @VessOnSecurity

Attending the third and last day of the ethics conference.

Before we begin I stop by one of the organizers' table and half-jokingly note that I hope there will be no racist remarks today.

"What racist remarks", asks the organizer. Oh, c'mon, you were sitting on the same row as me yesterday during THAT lecture.

OK, let's be precise. "Yesterday morning, during the first lecture", specify I.

Another attendee, standing besides me, immediately grasps who and what I am talking about and provides the name of the person. A name, which I never bothered remembering. Because I didn't want to.

"Oh, that one", says the organizer. "He's, you know, like this. Our mascot, sort of. A bit zany. He's like an old uncle, you know."

"He wasn't any better when he was younger", helpfully supplies another organizer. I don't doubt it.

"If he's ignored, he starts complaining that he's being repressed. So, you know, we had to invite him.", says organizer #1.

Dunno... When the Jewish people were gassed by the millions during WW2 - they were repressed. Not inviting an old, zany, anti-Semitic uncle to an ethics conference doesn't quite rate as being repressed in my book. But, hey, what do I know. I'm not an expert in this kind of thing.

Whatever. The conference begins. Immediately, I get the impression that I'm attending a different conference today. Apparently, all the kooks and crank talks were confined to the second day. Today it's a real conference on ethics.

While much of the research presented is outside my fields of interest, at least it's professional research, as you'd expect from a scientific conference.

My only complaint is that most of the speakers still don't have a clue how to present. Almost all of them read their papers in a monotonic voice, or don't have an accompanying presentation, or the presentation consists of walls of text on each slide, or a combination thereof.

The first lecture is hands down the best of those presented at the conference. The speaker actually talks to the audience, instead of reading under his nose. The presented ideas show deep thought, are quite interesting, and provocative.

The subject is "Can we teach AI to distinguish between good and evil?".

I really don't want to comment on it; it wouldn't do it justice. I got a copy from the author; I need to read it carefully and think hard about it.

The speaker is older than me, white hair and beard. I expect him to be a professor of something - maybe philosophy.

Since I've expressed interest in his work, he gives me his business card. His background is in... cinematography?!

Later I learn that he simply got interested in the subject of AI, wanted to learn more, and became a Ph.D. student. He's currently working on his thesis.

Wow.

The depth of his thought could easily put in his shirt pocket half of the room full of professors and doctors.

The next speaker talks about weak, strong, and super-strong AI. In reality, it's an overview of AI development. She isn't saying anything I didn't know already - but isn't saying anything wrong, either.

The next speaker is talking about ethical hacking. I kinda disagree with a bunch of things she says, although not strongly enough to object. I'm not sure if she's really confusing the issues or doesn't express clearly enough that she understands the difference, but...

She leaves the impression that when talking about "ethical hackers" she doesn't distinguish between (a) a person who is CEH-certified, (b) a hacker who is ethical, (c) a pentester and (d) a cyber-security specialist. After the lecture I privately explain her difference.

She also says that hackers implemented various programming languages like BASIC and Unix. This draws a grunt from somebody in the back rows, so I'm not the only one who is spotting the error.

Later, in a private conversation, I learn that she really doesn't know that Unix is an OS - not a programming language. She's too young to remember when Unix was popular (and Linux didn't exist), so she's excused.

She's also calling the early hackers "left libertarians and anarchists", which gets my hackles up, because I consider myself a libertarian and I would describe my political views as anything but "left".

Turns out it's a misunderstanding. She meant "left" as a noun - not as an adjective. She was talking about "the left, the libertarians, and the anarchists", so it's all good.

The next speaker is talking about "who owns the bug". I thought it would be about vulnerability disclosure - but she's talking about machine vision, pattern recognition, clearing an image of visual errors, that sort of thing. Apparently, she's confusing "bugs" with "errors".

The next speaker is a rather outspoken blond middle-aged lady who is talking about how the "robots are destroying contemporary journalism". She's specializing in public media (of course). A professor of this subject, actually.

Apparently, her problem is that AI is increasingly used in journalism to write articles and she's worrying that this would kill deep, thoughtful, investigative journalism and will replace it with mechanically generated cheap and shallow articles by the ton.

She's not exactly wrong but we aren't there yet. She has been impressed by how well AI can simulate music from the classic authors - close enough to fool experts that a lost work of a famous composer has been found. She thinks the same will happen with writing.

A student from the back row timidly objects that it's not quite the same. Music affects our emotions on a very low level and is much easier to simulate. Writing, with the exception of some trivial things like very short stories, is much harder.

It requires a plot, character development, etc. Just because AI can construct sentences using the same word frequency and forms of speech as Tolstoy doesn't mean that it can write "War and Peace". Maybe one day it will be able to but we're not quite there yet.

The media professor disagrees. The media professor is wrong. The timid student is right. I make a note to tell the student.

The media professor also seems to have taken half of her understanding of AI from SciFi, like Azimov's stories. She points out that some of his Laws of Robotics have already been broken by existing robots.

I regret to inform her that SciFi is not the real world and that Azimov's Laws were a metaphor to describe an ethical human and not really a blueprint followed by real-world roboticists today.

The next speaker is a law professor, talking about "words against neutrons". Initially I think that he's confusing neutrons with electrons but he's not. He's referring to the neutron radiation released during the Chernobyl nuclear incident and the attempts to cover up the truth.

Except that he isn't referring to the actual incident but to the TV series with the same name. I kindly advise him just like the media professor, to get his scientific information from scientific articles and not from works of fiction, no matter how realistic.

I also notice that the media professor, the law professor, and a bunch of people from the public use the word "algorithms" as the people of old used the word "magic" - to cover anything and everything they don't understand.

I inform him that "algorithms" is not the magic beans of computers, that they are just "recipes" and that even if he knew all the algorithms and source code of Google and Facebook, he still wouldn't know why exactly they behave the way they do.

He's a bit miffed, interrupts me, and says that he knows everything about the "black boxes of machine learning". I honestly doubt that he knows everything on that subject, but since he seems to know enough to grasp what I was referring to, I let it be.

A few more talks follow. They are totally uninteresting to me but I don't see anything wrong with them, either - they are just on subjects I don't know much about and that are outside my fields of interest. So, I don't remember them.

At some point of time comes the subject of computer games and their impact on children. You know how the media keeps harping that violent computer games cause violence, right?

Well, at least three specialists (psychologists) at the conference politely point out that this is bullshit.

One of them says that children understand perfectly well what is a computer game and what is the real world and that if any, the correlation between violent games and real-life violence is slightly negative, because the games help the children vent frustration harmlessly.

Another one has a whole talk on the subject. She's done a lot of research, including on her own two kids. Some of the results are blindingly obvious. Like, if a kid has a tendency of violence, the same kid is likely to engage in cyber-bullying too. Well, duh.

But there is less obvious stuff, too. She has studied how violent computer games affect children of various age groups. Apparently the age 10-12 is the most critical one.

Before children are just not that interested in such games, after that they have mostly formed their criteria of good and evil and the games don't impact them.

She says there is occasional negative impact (like, kids becoming more agitated and prone to violence) but it is very short-lived and immediately after the game.

Surprisingly (or not) it's not violent games that have such impact per se but competitive games where one team of players is pitted against another.

Some professor of psychology (I think) is talking about something I can't translate. It's something similar to morality (and this is how Google Translate translates it) but is neither morality, nor ethics.

I don't think that there's an exact English word for it. We do have words for "morality" and "ethics" that are different from it. The word in question is "нравственост".

Another speaker talks about digital literacy and how some study has shown that 18% of the Bulgarians are digitally literate. I ask what exactly does this mean. Like, when is a person digitally literate? When they can turn their computer on? Create a Facebook account? Install OS?

Somebody from the public (the media professor, I think) recites the EU's legal definition of the term. Which, like most things EU, makes zero sense.

The speaker objects that this is an old and obsolete definition, says that now there is a precise list of things that defines when a person is digitally literate and promises to send me a link to that list.

Some medical doctor talks about ADHD, how it is becoming increasingly widespread, how children are given Ritalin to combat it, and what the negative effects are.

This wraps up the conference, I think.