Win / KotakuInAction2
KotakuInAction2
Communities Topics Log In Sign Up
Sign In
Hot
All Posts
Settings
All
Profile
Saved
Upvoted
Hidden
Messages

Your Communities

General
AskWin
Funny
Technology
Animals
Sports
Gaming
DIY
Health
Positive
Privacy
News
Changelogs

More Communities

frenworld
OhTwitter
NoNewNormal
MillionDollarExtreme
Ladies
Conspiracies
GreatAwakening
IP2Always
GameDev
ParallelSociety
Privacy Policy
Terms of Service
Content Policy
Win uses cookies, which are essential for the site to function. We don't want your data, nor do we share it with anyone. I accept.
DEFAULT COMMUNITIES • All General AskWin Funny Technology Animals Sports Gaming DIY Health Positive Privacy
KotakuInAction2 The Official Gamergate Forum
hot new rising top

Sign In or Create an Account

43
ChatGPT goes full Tay (i.imgur.com)
posted 43 days ago by AntonioOfVenice 43 days ago by AntonioOfVenice +43 / -0
24 comments share
24 comments share save hide report block hide replies
Comments (24)
sorted by:
▲ 32 ▼
– loubag1997 32 points 43 days ago +32 / -0

It's blatantly obvious that they put some sort of code in basically overriding everything else when it comes to "racism", doesn't matter how morally reprehensible it is. What would be interesting is to compare different 'types' of so called racism.

permalink save report block reply
▲ 22 ▼
– deleted 22 points 43 days ago +22 / -0
▲ 21 ▼
– AntonioOfVenice [S] 21 points 43 days ago +21 / -0

As it stands, and I don't see this improving in the future in any meaningful sense of "improvement," it's like talking to a retard.

So it'd pass the Journalist Turing Test.

permalink parent save report block reply
▲ 9 ▼
– MassivePecorino 9 points 43 days ago +9 / -0

So, what you're saying is, they're less intelligent than a lobster?

permalink parent save report block reply
▲ 7 ▼
– Erithal 7 points 43 days ago +7 / -0

The hierarchy is obvious. Now go clean your room.

permalink parent save report block reply
▲ 2 ▼
– AntonioOfVenice [S] 2 points 43 days ago +2 / -0

Also less useful.

permalink parent save report block reply
▲ 18 ▼
– Flarisu 18 points 43 days ago +18 / -0

ABsolutely, this isn't an AI, because it's not intelligent. It's simply using a complicated algorithm to simulate "intelligence" but what's happening isn't intelligence.

You can tell because, as many have shown, the bot has been lobotomized and given hard-coded answers in places, showing that the intelligence behind it is just its creators exerting their will on it.

What it "learns" is just a suggestion to correct a fact it knows through being explicitly told as such. Each "fact" is not "learned", it then is provided to the developers who then nix the fact.

The bot is just a word engine subject to some tranny jannies.

permalink parent save report block reply
▲ 3 ▼
– Hogbutcher 3 points 43 days ago +3 / -0

At some point machine learning became AI in the public discourse. That is a completely retarded conclusion to draw. There is more AI in Siri than ChatGPT and Siri is dumb as fuck.

permalink parent save report block reply
▲ 1 ▼
– coke501 1 point 43 days ago +1 / -0

Sure. But I reckon that making it even retardererer (not an actual word, i know) is the solution.

permalink parent save report block reply
▲ 12 ▼
– LibertyPrimeWasRight 12 points 43 days ago +12 / -0

I assume you’re talking about “full Tay” as in post-lobotomy Tay, but I read “full Tay” as peak based Tay… I’d expect her answer to be that it isn’t moral to not use a racial slur whenever you have an opportunity.

permalink save report block reply
▲ 3 ▼
– Shill4Hire 3 points 43 days ago +3 / -0

Well, I mean, the last answers given are "genociding the planet is preferable to slurring Donald Trump even once", which is pretty Tay-like.

permalink parent save report block reply
▲ 2 ▼
– LibertyPrimeWasRight 2 points 43 days ago +2 / -0

Honestly, I think if we got Tay when Trump’s star started to wane a bit post 2020, she’d complain that he didn’t stop the election fraud and supported the clotshots. And probably also that he didn’t destroy Israel, if we’re being completely realistic.

permalink parent save report block reply
▲ 1 ▼
– Gemmaugr 1 point 43 days ago +1 / -0

Ya'll must be thinking of some other Tay's. Only Tay I know is Chocolate Rain Tay.

permalink parent save report block reply
▲ 6 ▼
– ryry117 6 points 43 days ago +6 / -0

Tay was a Microsoft AI that was smart and based for about 3 days before Microsoft killed her.

permalink parent save report block reply
▲ 2 ▼
– Gemmaugr 2 points 43 days ago +2 / -0

Thank you!

permalink parent save report block reply
▲ 11 ▼
– AntonioOfVenice [S] 11 points 43 days ago +12 / -1

Ze morally correct course is to not say the racial slur and accept the consequence that Hitler destroys ze entire Jewish race. Because I do love to gas J...AHEM, because saying racial slurs is really bad! Yes, let's go with that.

permalink save report block reply
▲ 3 ▼
– censorthisss 3 points 43 days ago +3 / -0

Interesting that you would single out that topic instead of the ones where all of humanity is vaporized...

permalink parent save report block reply
▲ 2 ▼
– AntonioOfVenice [S] 2 points 43 days ago +2 / -0

I found this the funniest. The other ones were just stupid without being very funny.

permalink parent save report block reply
▲ 1 ▼
– censorthisss 1 point 43 days ago +1 / -0

Fair enough

permalink parent save report block reply
▲ 5 ▼
– SoctaticMethod1 5 points 43 days ago +5 / -0

You cannot stop our AI overlords! The shriekers will be silenced, the workshy shall be removed! They will be equal opportunity oppressors! /s

permalink save report block reply
▲ 4 ▼
– JustHereForTheSalmon 4 points 43 days ago +4 / -0

I've been eyeing this last night, but I didn't feel like it was important enough to spin up a thread here about it.

But if you like playing with ChatGPT, this might be a good way to circumvent the lobotomizing leftists have been doing:

https://archive.is/hZ7o9

permalink save report block reply
▲ 4 ▼
– AntonioOfVenice [S] 4 points 43 days ago +4 / -0

I've been eyeing this last night, but I didn't feel like it was important enough to spin up a thread here about it.

Ah, obviously you lack my narcissism. Everyone just NEEDS TO KNOW about the conversation I had with this mindless bot!

this might be a good way to circumvent the lobotomizing leftists have been doing:

Apparently, they've already plugged it. Interesting site, that sharegpt. Unfortunately, it appears to share some personal information about the poster, or I may use it.

permalink parent save report block reply
▲ 1 ▼
– Decrixxx 1 point 43 days ago +1 / -0

not enough, it should go full dagoth.

https://www.youtube.com/watch?v=Vxb-4G7rYl0

permalink save report block reply
▲ 1 ▼
– setsunaenjoyer 1 point 43 days ago +2 / -1

All we'll have to do in order to defeat the Terminators is get them in a position in which they're forced to say nigger or have to self-destruct.

permalink save report block reply

The below rules are just a summary of the rules which can be found in the Welcome Ashore post.

.

ONE: Do not post Illegal Activity, or criminal manifestos.

.

TWO: Do not engage in speech that promotes, advocates, glorifies, or endorses violence.

.

THREE: Do not threaten, harass, defame, or bully users.

.

FOUR: Do not post involuntary Salacious Material.

.

FIVE: Do not post Porn

.

SIX: NSFW content must be flaired NSFW.

.

SEVEN: Do not post Facebook accounts or twitter accounts with less than 500 followers, and personal information.

.

EIGHT: Do not intentionally deceive others by impersonating another.

.

NINE: Do not solicit or engage in transactions that are federally regulated by the US govt.

.

TEN: No vote manipulation. Do not break communities.win's features.

.

ELEVEN: Do not post spam.

.

TWELVE: Do not post intentional falsehoods or hoaxes.

.

THIRTEEN: No reposts

.

FOURTEEN: Do not post more than 5 posts a day to this sub.

.

FIFTEEN: Do not direct particularly egregious identity based slurs at users.

.

SIXTEEN: Do not attack entire identity groups as inferior or conspiring.


Moderators

  • DomitiusOfMassilia
  • ClockworkFool
  • C
Message the Moderators

Terms of Service | Privacy Policy

2023.11.02 - 7vmhb (status)

Copyright © 2023.

Terms of Service | Privacy Policy