State of Internet Replies
I say, since weβll soon have to solve the problem of terrible AI-generated replies,
how about we solve terrible human-generated replies too, while weβre at it?
I say, since weβll soon have to solve the problem of terrible AI-generated replies,
how about we solve terrible human-generated replies too, while weβre at it?
π13π5β€3π2π1
This media is not supported in your browser
VIEW IN TELEGRAM
Rugby journalist in Australia forgets to remove the ChatGPT prompts before posting article
π33π7π5π3β€2π2π2π₯1
Adobe and Shutterstock, Microsoft, and other big companies working to lower risks of using their AI tools by promising protection in case of copyright lawsuits
Microsoft Article
Shutterstock Article
Adobe Article
Microsoft Article
Shutterstock Article
Adobe Article
π₯26π10β€7π7π4π3π―2π€£2
This media is not supported in your browser
VIEW IN TELEGRAM
ChatGPT now deleting messages if flagged by moderation
And itβs flagging everything.
Today it deletes your messages,
Tomorrow it deletes you.
And itβs flagging everything.
Today it deletes your messages,
Tomorrow it deletes you.
π€¬37π«‘20β€14π11π8π¦7π±2
What will future AIs be able to do that todayβs canβt?
Got one idea, bro.
Verbal intelligence clearly leads to many general, reusable skills, far more than what youβd could say is βjust playing with wordsβ, as LLMs have already shown.
Likewise, strong evidence that high visuospatial skills also lead to a huge range of general reusable skills, far beyond what you could say is βjust playing with shapesβ.
Personally, often use mental imagery as a mental scratchpad, to visualize multiple distinct things in cases where words otherwise fail to distinguish them.
One would guess that this could explains why those lacking on visuospatial skills, but high on verbal skills, the wordcels, are apparently so often falling into obvious equivocation traps.
With their words alone failing to distinguish a single thing from many, theyβre left wandering blindfolded in mazes of their own equivocation lies.
When BroGPT?
Got one idea, bro.
Verbal intelligence clearly leads to many general, reusable skills, far more than what youβd could say is βjust playing with wordsβ, as LLMs have already shown.
Likewise, strong evidence that high visuospatial skills also lead to a huge range of general reusable skills, far beyond what you could say is βjust playing with shapesβ.
Personally, often use mental imagery as a mental scratchpad, to visualize multiple distinct things in cases where words otherwise fail to distinguish them.
One would guess that this could explains why those lacking on visuospatial skills, but high on verbal skills, the wordcels, are apparently so often falling into obvious equivocation traps.
With their words alone failing to distinguish a single thing from many, theyβre left wandering blindfolded in mazes of their own equivocation lies.
When BroGPT?
πΏ15β€7π5π3π1
The fraud of Originality.ai, and similar AI detectors
βOriginality.ai's only proposed solution is for me to download their Google Docs plugin and purchase credits to run reports proving the documentsβ authenticity.β
Students and anyone else happily handing over their Google docs history is insane.
βOriginality.ai's only proposed solution is for me to download their Google Docs plugin and purchase credits to run reports proving the documentsβ authenticity.β
Students and anyone else happily handing over their Google docs history is insane.
π€¬19β€3π3π1π₯1
Chat GPT
If you could make just one demand, either maximum AI transparency or maximum control over AI, which would it be?
AI Transparency vs. AI Control Poll Results
π€―8β€5π4π‘2π1π1
Chat GPT
AI Transparency vs. AI Control Poll Results
So, would you rather live in a world with AI that is 100% TRANSPARENT that it is deleting all of your comments and making you say whatever it likes, or an AI where you can CONTROL whether or not it deletes your comments and forces you to say things.
Anonymous Poll
40%
Total AI Transparency
43%
Total AI Control
17%
Show Results
π₯13π8π4β€3π3π2π2π1
Majority of βAI girlfriendβ app users are actually female - A16z partner
βIt mimics fan fiction, where ~80% of readers are women.β
Exactly what weβve been saying.
Guys, if you didnβt realize that a huge percent of women are into romance novels, lit-erotica, and similar text-based sexual fantasies, and into it far more than guys, then bro you donβt know the first thing about women. (Goes for most women too, who tend to know even less about other women.)
AI bfs gonna steal your girl.
Link
βIt mimics fan fiction, where ~80% of readers are women.β
Exactly what weβve been saying.
Guys, if you didnβt realize that a huge percent of women are into romance novels, lit-erotica, and similar text-based sexual fantasies, and into it far more than guys, then bro you donβt know the first thing about women. (Goes for most women too, who tend to know even less about other women.)
AI bfs gonna steal your girl.
Link
π€―19π€£13β€7π3π2π₯1π±1π1