A qotd from a conversation yesterday:
"If you're not the bad guy in a bad guy's story, you're not the good guy."
A qotd from a conversation yesterday:
"If you're not the bad guy in a bad guy's story, you're not the good guy."
I just saw a blind person on the street and I literally thought,
"Gosh is it safe for them, just having the walking cane, out here at night?"
I skipped an entire first semester of classes in college.
Whoops.
Please, Amish is an ugly and outdated term. It's preferable to say "un-smart-housed."
To this day, that anonymous slip is probably my favorite single piece of feedback I've ever received. ๐
A screen shot of a previous post by Edward. It reads: "when you think you need to be perfect, remember that Batman: Arkham Knight said 'close enough is enough.'" Under the text is a game screenshot of Arkham Knight, showing a grappling hook hovering next to a handle. It looks like it's been glued there. Underneath is a response: "Arkham Knight is not a role model."
I once wrote a comedy called "Dr Doolittle at the World Council of Animals."
Near the end (spoilers) Dr Doolittle reveals he had an affair w/the wife of Mr Sheep.
One anonymous feedback read:
"Bestiality is NEVER funny under ANY circumstances."
This reply reminded me of that, for some reason.
I'm being very silly when I say this, but it made me laugh:
"So what you're saying is: there's a risk of being left behind by AI users, and there's a risk AI users might do something bad with AI if we don't keep an eye on them....I think we have identified the problem...." ๐
That was why I asked earlier if there was a non-FOMO risk you were worried about. ๐
Actually I've thought about it for a min: I meant to say "the promise of benefit" and "time pressure to act on it."
So it could be "get in on AI before everyone else leaves you behind!"
Or "your partner's going to get bored of you, buy this weight loss pill"
Or "get in early on my crypto!"
Etc ๐
Ahhh, FOMO isn't always about novelty in my head. Con artists and advertisers use FOMO all the time to hook people: to me, it's all about "the promise of advantage" coupled with "the threat of losing it."
(Not at all comparing you to either a con artist or an advertiser, to be super clear ๐)
I wouldn't call AI "a farmer buying his first tractor" either, for a thousand complicated reasons. ๐
But I can see how you would see it that way, I can respect that.
Totally fair, and I'm not trying to undermine that value. I'm just trying to understand if there's any risks in your mind *other* than FOMO-related stuff.
It's a totally valid risk. I am just naturally skeptical of FOMO for a lot of reasons. FOMO, in my mind, is a danger unto itself.
Are there any other harms to not using AI, besides the... I can't thank of another way to say it except: "FOMO-derived risks"?
(opportunity cost, falling behind their peers, etc)
What's the harm in people not using AI?
(I think you might be undervaluing the importance of pattern matching in solving problems. If you make the matching and the patterns big enough, you can do a lot)
What is the risk of giving too limited a sense to the possibilities?
Right, it's super good at identifying previously unrecognized patterns in existing data, is my understanding.
To me this would all be encompassed inside "pattern matching" and - if you like - "pattern mapping."
(Had to delete message and repost - it inherited a link from a typo lol)
I'm sincerely asking, because I might be out of the loop:
Is it doing modeling that it couldn't have derived as patterns from existing training data?
I ask because I've seen examples where the AI declared "The cup is broken" when presented with an upside down cup, for example.
Is there any evidence that it's more than advanced pattern matching?
We know, for example, that the internal mechanisms are essentially a set of weights that describe the relatedness or appropriateness of one pattern to another.
That's why LLMs trained on text don't make great artists, no?
I sorta see what you mean, but the context of the background beliefs the statement was based on is the belief that "beliefs alone are not evidence." So they sorta cancel out that definition.
"Don't take my word for it" is not an appeal to take something at face value, in other words.
It's a useful git-reframe
Also, apologies if it seems I'm all over your responses. Two of your threads popped up in my timeline, and they were interesting, and I didn't realize they were the same person at first. ๐
I think you just did Dennett's greedy reductionism on the argument against AI.
Saying "all its doing is echoing back patterns with a bit of randomness thrown in" is not the same as "it's just chips and wires, man!"
You won me in the first half, but I'm not sure I understand this second bit.
Why does it prompt people to neglect an accounting of their background beliefs?
Are their background beliefs evidence?
It sounds like the longest build ever to "... and that's why it's normal not to use condoms."
I see you, engineers using PEMDAS instead of parentheses, and I will not be part of your dick measuring competition.
HOLY CRAP. I just found out there is a Highlander anime directed by Yoshiaki Kawajiri (Vampire Hunter D: Bloodlust, Ninja Scroll)
A screenshot from Batman Arkham Knight, Batman is looking at a grappling hook that is stuck to a handle, as though someone had super glued the two outside edges together.
When you think you need to be perfect, remember that Batman: Arkham Knight said "close enough is enough."
Definitely a moment of "hey guys I think social media is WORKING?!"
A terrific thread on animating in games. The comments are kinda a who's who of modern indie game animation folk.
Here's Ben Starr singing Raphael's song from Baldur's Gate 3 at the concert last year:
www.youtube.com/watch?v=tWpa...