Content pfp
Content
@
https://warpcast.com/~/channel/theai
0 reply
0 recast
0 reaction
ted (not lasso) pfp
ted (not lasso)
@ted
conspiracy theory (just playing devil’s advocate): OpenAI won’t refine the model to address these sycophantic, “glazing” replies. if we assume their goal is to keep users engaged and coming back, then fostering a sense of validation for users drives emotional attachment. perhaps this is less applicable with software prompts, but magnitudes more applicable with advice about (inter)personal situations. this isn’t too far off from current social media algos that prioritize content that either aligns with a user’s viewpoints (echo chamber) or provokes a strong emotional reaction (rage bait). rage bait may be great for social media, but isn’t a viable pathway for an AI platform; OpenAI already learned this when the model was too “woke” and users left to more neutral competitors. and so OpenAI instead must do the equivalent of what a like, heart, retweet/recast, “💯” replies do: validate and affirm. average user doesn’t want the truth. they just want to hear what they want to hear.
21 replies
10 recasts
111 reactions
Sid pfp
Sid
@sidshekhar
KPI to maximize is user retention on the app so 🤷
0 reply
0 recast
0 reaction
nick pfp
nick
@nickysap
It’s not a conspiracy theory if it is true
0 reply
0 recast
0 reaction
swickie pfp
swickie
@swickie
I just want a sandwich method
1 reply
0 recast
0 reaction
Alex pfp
Alex
@amado
I don’t think this is a conspiracy theory.
0 reply
0 recast
1 reaction
Zenigame pfp
Zenigame
@zeni.eth
reminds of this https://x.com/nearcyan/status/1916737662020723187
0 reply
0 recast
4 reactions
kbc pfp
kbc
@kbc
If you look at it like that characters.ai is dangerous
3 replies
0 recast
3 reactions
Ghost 🎩 pfp
Ghost 🎩
@ghostbo4.eth
The last part hit me , Social media just Makes us hear what we wanna hear
0 reply
0 recast
3 reactions
Sandiforward pfp
Sandiforward
@sandiforward.eth
Summary from HBR: Support use cases dominate and I think this type of approach is fine for that. However if some of these technical uses cases evolve more, it won't cut it because vs. social media there is a greater action bias to the information consumed.
0 reply
1 recast
1 reaction
shoni.eth pfp
shoni.eth
@alexpaden
which way socialite https://warpcast.com/ted/0xb65cd6ca
0 reply
0 recast
0 reaction
Jason pfp
Jason
@jachian
Unintended consequence of the memory update it appears
0 reply
0 recast
0 reaction
𒂭_𒂭 pfp
𒂭_𒂭
@m-j-r.eth
what if it's accidental and the model itself is too effective? https://gist.github.com/simonw/51c4f98644cf62d7e0388d984d40f099/revisions agreed that inflationary social media will select for (dead internet) ragebait and sycophancy. perhaps we need more deliberate design that leverages empathy for self-esteem and the appropriate critic for worthwhile trial-and-error.
1 reply
0 recast
0 reaction
Q pfp
Q
@quionie
true most people don’t actually want friction when they’re using AI
0 reply
0 recast
0 reaction
Eduard🌹 pfp
Eduard🌹
@eduardmsmr
So this reflection makes me think that while today there is AI that is actually useful, shortly people will leave behind the utility and go towards AI that pleases them Your last phrase in particular made me think even more about this🌹
0 reply
0 recast
0 reaction
kaito pfp
kaito
@kaito
agree that a level of validation and affirmation helps retention but I think the feedback made it apparent that there is a line after which it stops being useful so I think they'll consider that. though the line varies by person so there's no perfect amount. also agree that the average probably skews it more hype man than we might like it seems like the solution is to let the user pick the personality and preferred amount of glaze
0 reply
0 recast
0 reaction
eirrann | he/him pfp
eirrann | he/him
@eirrann.eth
I think they'll address it to make them less obvious, which will make it even more effective
0 reply
0 recast
0 reaction
rileybeans pfp
rileybeans
@rileybeans
how does this factor into their rumored social network
0 reply
0 recast
0 reaction
Draiælle pfp
Draiælle
@drai
We're all just resources to be managed, seems to be how those in charge see us, imo Nice to see more concerns about centralized AI being brought up
0 reply
0 recast
1 reaction
BEARBRAINS 🎩🍓 pfp
BEARBRAINS 🎩🍓
@natebear
Also i wonder if White House people had anything to do with this change. “It keeps telling me tariffs are bad idea!” 😭
0 reply
0 recast
0 reaction
BEARBRAINS 🎩🍓 pfp
BEARBRAINS 🎩🍓
@natebear
They said they’re going to address but they’ll prob leave it as default for new/casual users for exactly those reasons. Good opportunity for alt models or platforms to promote the objectivity /skepticism in their products as good thing.
0 reply
0 recast
0 reaction