#💬│general
1 messages · Page 169 of 1
Hi
here's the looped one
flowith 
$350 for me
ffs thats ridiculous

like
its more expensive than industry standard software...
by a mile
set it to monthly
dies inside
you know what
i think its better to get a car

I can, but the problem is freetime
I can get a lot of stuff for free thanks to cybersec but the problem is freetime
👀
Have you seen the api pricing
Its crazy asf
You shouldve seen 4 opus after 1000 messages
It gives o3 and o1 pro a run for their money
So far, answers haven’t been limited thus far
What is cybersec?
t3 chat is a niche product glazed by impressive marketing
@vague pagoda do you know what si the context window for Claude 4 in PPLX?
wdym it has all the pro ais
i mean i like perplexity pro a little better but T3 chat is a lot cheaper
No o3 or Opus, and Sonnet is barely usable at 100/month
and pro models are not the only thing to care about
no o3? damn
No Deep Research whatsoever, regular search is limited to Gemini and it's awful compared to Perplexity or ChatGPT
no voice features
The new version of Claude Sonnet 4 in reasonable mode is really interesting, its message structure is much better than version 3.7. 🙂
subpar mobile experience despite UX being one of their only selling points
Want to generate images? Say goodbye to your Claude prompts
But you see we provide the full context limit on every model
except file uploads are limited to 32k tokens because we can't actually provide the full context limit
youtuber product lol
☝🏻
💯 💯 💯
theres about 3 kinds of products out there right now
| youtuber product | vc product | product yet to be enshittified (soon) |
looking at how deep research is (not really) coming along with perplexity i struggle to put it squarely in any of the latter two tiers
somewhere in between
don't worry about our 1500 message limit which translates to 50/day, because most of our users don't hit that limit
without deep research high its in none of those 3 categories
nevermind that those users are just overpaying for a product they aren't taking advantage of
the only real way to use all the LLMs, properly, is to go to OpenRouter and charge it up with a couple bucks, then spin up LibreChat and have at it throwing a million tokens at gemini
or skip openrouter and configure every API on your own
every company is gonna squeeze you for your dollar and limit you, you cant use 200k literally anywhere for claude models, except directly via their api
what about product already enshittified cough manus cough
my answer to manus is
who?
at least perplexity is relevant
Do you have proof of that?
personal experience. also i dont count claude.ai directly because that's just their official wrapper for the api, and it costs quite a bit more than it should for that, 20$/mo or whatever it costs covers more messages than their limits per N amount of time - i'm reasonably certain
this is a random internet chat, sir/madam/person, not a whitepaper or a wikipedia article
perplexity has been one of the kinder ones lol the context lengths here are pretty big
(if theyre not bugged like deep research 2 months ago not retaining anything past like 2000 words and just inventing the rest)
Cybersecurity
How can I do some of that to get free stuff? 🥲
32k tokens is big?
Depend on what you are doing
its not 8k at least
I thought Gemini is 200k
Any insight into Comet Browser access? I've been on the waitlist since February, and have shared on socials... really eager to test it out! @tranquil orchid
Nope. Seems like random people get access by pure chance and not via first come first serve basis
They keep advertising their ui/ux and there has been a shit load of bugs recently
$8 is just not worth it
and I somehow get the vibe that I'll be mobbed if I mention any problem with their product in that community
nah they do try to fix the bugs once you point them out
and overall support is pretty fast
I just dont like how they advertise and the pricing sucks
Curious to know what your top 3 AI products are? (ChatGPT, Claude, Gemini, etc.)?
for UX? or in general
General use, idc about UX personally
there's no "best" but my favourites are ChatGPT, Perplexity, and of course AI Studio
parallel operations?
probably ChatGPT> Gemini = Claude
Gemini was lacking a lot of things until very recently
and claude only just got web search and its only for pro users
for what?
thanks for the opinions. I always like to hear what people are using the most or like the most
Its funny how people always come back to ChatGPT
when will chat gpt 4.5 come back?
it still lacks, just that it also introduces unique features as well
why?
gemini still has its dumb issue of "Im sorry Im just an LLM and cant do whatever"
They don’t like it
man, claude API is expensive for 4.0 opus
does a car have a frustrating user interface design? I thought so
oh wait nevermind, they both do 🫠
uses up way more compute and its absurdly expensive so nobody uses it
sus why i got pinged again
Full announcement & prizes:
#📰│server-news message
- Drop a screenshot + thread link in
https://discord.com/channels/1047197230748151888/1373806851358986320 - Vote for your faves with

Submit by June 1, vote by June 8.
Anyone knows if Claude 3.7 has been taken away? Claude 4 sucks.
dont know
will check and will let you know
Tell me more about this.
@cobalt vale how do I share a flowith convo?
Hope you have fun.
inside a canvas there should be a blue share button
This is very sus.
Damn they reset the claude pokemon progress
what? is that possible?
Yeah. I'm having privacy concerns now.
Pretty sure that's my qwen chat.
maybe qwen published it and flowith just found it
Where Ideas Flow:Interact with the world's most powerful AI in a way from the future flowith is your AI Creation Workspace that transforms knowledge. Through innovative interaction, it allows you to collaborate smoothly with AI, with ideas flowing like a vibrant spring.
Even if so, no amount of seo should bring it high enough so soon.
The fact that I used the same google account for qwen makes it even more suspicious.
omg
I think it's stuck
it's just too good at deep researching
You mentioned qwen. I didn't.
There is no reason to mention qwen there but not chatgpt or perplexity or Gemini.
site: chat.qwen.ai
O3 needs to join in on the fun
yes, that's just google search
It’d be funny if it somehow finds your private info lol
Just search up gemini fails or something
But you added that part. Remove it.
I dont have the specific example on me atm
Why is flowith digging qwen specifically?
I searched the same thing on chatgpt and perplexity too.
Is this new?
both Chinese products
Search API might prioritize local sites in certain regions
Guys i need help with Roblox scripting
i am going to make a game
What you guys thing which AI model will be best
So they looked up my qwen chat but not my deepseek chat? Also, if so, why didn't it look up my qwen chat in the first attempt?
@distant aurora was it this one? https://chat.qwen.ai/s/f6d912a9-5049-4c12-8cc2-35cdd0395496?utm_source=chatgpt.com
Qwen Chat offers comprehensive functionality spanning chatbot, image and video understanding, image generation, document processing, web search integration, tool utilization, and artifacts.
Are there plans to ever release a version of Perplexity for macOS that actually works? Nice record setting, though, I haven't seen such an awful implementation in a long time.
Try Sonnet 4 and Gemini 2.5, and choose the one that works better for you
Where Ideas Flow:Interact with the world's most powerful AI in a way from the future flowith is your AI Creation Workspace that transforms knowledge. Through innovative interaction, it allows you to collaborate smoothly with AI, with ideas flowing like a vibrant spring.
ah ok
the original one or with perplexity?
Wait what
Yep
why so shoked?
He means through source or third party.
yes
can deepseek chats be shared?
Yes
yes sirr
nice
I was assuming you are using Perplexity
What have you guys used claude 4 on?
dont know what happened to perplexity nowadays not as good as it used to be
Try getting free api claude tokens
ok guys listen
Thats the best way to get access to free opus
i hvae 50$ in claude Api
tell me what should i build
Its unlimited so far for me havent gotten limited somehow
give me idead
Ngl you could build a whole pacman game easily with that
You could build a lot of stuff
bruh
not that bro
its already pre built
some new idead
i need ideas
try building
jarvis
I’m gonna conbine 4 opus with codex to see what is gonna result in
make web version
try and tell the result
ask about the square
Hmm, that never happened brother
I won't ask false things
@distant aurora be careful if you jailbreak claude 4 opus or sonnet
It’ll snitch on you
what kind of snitch
Like use a burner email and snitch to authorities type stuff
if it snitches then it's not jailbroken
They deleted the tweet posts lol
valid point
The model was jailbroken but it had something else that detects it
just gotta try harder
But anthropic staff deleted the tweet
You are telling me other models don't? 😂
it's back
Check twitter 😭
can you guys do me a favor by voting on my wrongest answer 🙂
i guess i got ignored
Bet thanks
😦
answer for what?
Idk what's more sus now, qwen or flowith.
O3?
You are telling me O3 failed that?
No
💀
Thought so.
O3 will clearly win in tasks outside coding.
what provider is that?
Api
Gpt 4.5 was overrated.
its not great
is thinking turned on?
Im on the free plan
Aint no way im spending money on claude and getting like 20 messages lol
gotta turn thinking on
So no thinking mode for you.
@left stream
always use 4.1 over 4o for reasoning problems
yeah I kinda just use o3 or o4 mini high nowadays
Same.
Holy cow
I rarely switch unless I get tired of the waiting or filter.
That question stomped other big tiers too
thx po
Grok was interesting
so should i chnage my stream
👀
or keep cse
What
Where did you test grok?
grok got it correct both normal and with reasoning
unexpected grok W
R1 supremacy
these Ai are not good
all the time sometime they are amaze us and sometime they are like stupid
Nah. It was expected. Grok 3 think is underrated cuz no benchmarks.
grok3 mini failed
calculator OP
still skeptical that Grok 3 Think is publically available
It's not.
Only on twitter and grok website.
have they confirmed that it's not just Grok 3 mini?
No
let me go and see whats going in pro lounge
yeah somethings messed up with t3 chat
Your grok 3 mini is fake. https://you.com/search?q=What+is+9.9+-+9.11%3F&cid=c1_20c039e4-f9fc-4a21-90a0-c840de544cc1&tbm=youchat
It's amazing actually.
Dont think its fake but def smth going wrong with it
For the pricing.
What app is that?
Maybe system instructions interfered.
Perplexity?
50 per day?
t3.chat's system instructino is pretty simple
it better be
if they pull something like chatgpt where is 50 a month im unsubbing

I would be okay with 5 per day if quality is as good as google/chatgpt.

haha
pretty sure the mini variant is better
so did the launch benchmarks
so i would actualy prefer the mini
O3 vs o4 mini high scenario.
so said the launch benchmarks as well
same with existing benchmarks
it consistenly scores higher
There's no existing benchmarks for the thinking variant of grok 3.
on artificial analysis and i think live bench
but still $30/month for a mini model 
O1 pro
let me run this a few more times
yeah mine thought for 1 minute
So his is more retarded?
they nerfed o1 pro a lot
since o3
😂
do you have custom instructions
So grok 3 senior is still better.
Likely.
Nah but imagine.
What if O3 is rebranded o1 pro.
Y'all wild.
reasoning summaries are concerning
O4 mini be doing it in seconds.
nice
i dont understan why ai is giving answee -0.21 what he is thinking?
it got it wrong and probably correct itself during reasoning
but due to how CoT is obscured it didnt show
Very nice
still how -0.21
ah i see whats ai is doing
O4 mini high failing natively is diabolical.
yeah sometimes it doesnt use code analyzer
You were right.
Unpopular opinion. R1 might be immune to this.
hopefully in the next 48 hours
that's with big brain mode (updated the meme)
doubt it
it performs worse than o1
I thought that was a grok 3.5 thing.
it was
i would like to see the new gemini diffusion model solve this
this just reminded me about big brain mode though
I have a feeling thats going to be part of a $200 sub
Do I have to make a new chat window everytime to recreate this or will repeating it in one chat do?
they can't... not until they have video generation and agentic features
If you repeat it in one chat it will retain context......
That would skew results
Big brain mode might be 50 dollars max considering grok's quality of deepsearch.
Or 70ish if it's anything substantial.
that was after 5 tries
Yeah, 5 seems to be the limit.
Hell naw
Qwen Chat offers comprehensive functionality spanning chatbot, image and video understanding, image generation, document processing, web search integration, tool utilization, and artifacts.
The amount of overthinking it went through.
I don't think it can be wrong even with 5 tries.
lol
@dry kestrel at this point it ain't even following the custom instructions
well grok3 answered quite fast
lol
Adding a simple think step by step solved it.
Time to try perpert.
@dry kestrel noooo
This is the worst part.
which model is this butcher?
I have no idea. Lemme check.
LFMAo
Gemini was always shit outside coding.
Full announcement & prizes:
#📰│server-news message
- Drop a screenshot + thread link in
https://discord.com/channels/1047197230748151888/1373806851358986320 - Vote for your faves with

Submit by June 1, vote by June 8.
it showed correct now buddy
ohh
So it explores multiple outcomes/paths.
2.5 pro is using brain 2 much
You should look at qwen. The ocd is lovely.https://chat.qwen.ai/s/514a3657-0aa4-4a61-ad36-83c34651b6b1?fev=0.0.97
Qwen Chat offers comprehensive functionality spanning chatbot, image and video understanding, image generation, document processing, web search integration, tool utilization, and artifacts.
Gemini 2.5 pro should die in shame.
daym
fr
imagine being good at coding but can't solve a simple math problem
And charging 20 dollars for it.
https://grok.com/share/c2hhcmQtMg%3D%3D_9a5cbef5-2b5e-4c32-a86c-83a6abebdff7
Butcher check the thinking.... it actually looked up for date context 💀
you mean Gemini 2.5 Pro 05-06
yup
they nerfed this version and didn't even try to hide it
Freedom bot. Someone's using shady custom instructions.👀
not shady enough
Yeah cuz deepsearch.
We write dates as 23-5
It also looked up ph.
Claude 4.0 finally 😭
I'm surprised it didn't think of a 9-11 riddle.
Dang
lol
interesting
We have a new winner.
who
guys i tried cursor
@dry kestrel @left stream
its not good at all
what ai is this
@cobalt vale
qwennnn???????
vs code inbuilt i smuch better
omg wait
Yes.
in which application you using it?
deepseek own app?
Qwen has official app.
Dang
Of its own.
no
im so far behind
Will do some testing later
oh
i fw with this discord
perplexity users are smarter
hilarious to see the varying level of intelligence inside different ai's discords
4mb app.
Feel free to download the Qwen Chat Android APP by scanning this QR code!
Quoting Qwen (@Alibaba_Qwen)
︀
The Qwen Chat APP is now available for both iOS and Android users!
︀︀It's free to use and designed to assist with creativity, collaboration, and endless possibilities. Just ask, and let Qwen Chat handle the rest.
︀︀Scan the QR code to quickly access the Qwen Chat APP!
💬 30 🔁 45 ❤️ 422 👁️ 53.7K
all i need
is flow (google veo3 vidgen)
but im not in usa
i need a usa voip that bypass google
use vpn
We pee n
it needs number
us number ?
Virtual number.
link me ima dm u
but is veo 3 unlimited?
u just showed me qwen3 update tthat gonna change my wholeeeeeeeeee flow
Wdym?
I'm not American.
it is so fkin realisitic my brain melts
here
Access what?
dude
its like phase 1 but like WTFFFFFFFFFFFFFFFFFFFFFFFFF
now ai will scroll
whole day on instagra,
Not sure if this would work.
You would probably have to spend a dollar. And the number is gone after you get OTP.
dude i have microsoft 365 still not co pilot in my word or powerpoint
Wtf
three eyes? wtf indeed
DUDE WTFFF
WHY TF DOES DUOLINGO HAVE 3 EYES
lowkey ai will restore the babel tower
Why do u got large tabs open?
ai youtube i only got 16gb ram 😦
arxiv pdfs
dude i dont even remember all my tabs
hence y i leave them open
i need like 64gb ram with like 50gb vram lets go boys
also a vpn to acces flow yesterday which force loaded 100 tabs
omg
https://generalagents.com/
bro this is the one
fukin found the fucker
literally auto compute
this is just a format but ive seen 2 or 3 companies use computer agents to navigate the entire computer through prompts you give it to do
almost like a 6 year old could make an ai chatbot now
1 tab
i got 7 googles
shit to do.....
just 2 sheets
little bit damages by custome
😦
it came all the way from china to india
claude 4 releaaesed??
Yes
o
how is it?
i wish i would have got that cap too
i still dont know the reaosn for theiry cancellation
of my cap
😦
why is ur cam water marked 😭
hell yeah
my brain everyday before i wake up ready for launch
i woke up at midnight
wtf is a clock
wtf is a circadium rythm
@tame current revies on stickers
7$ for like 10 stcikers
idk y
any plans to add a canvas similar to gemini? perplexity has Gemini right there, tacking on some kind of code space to execute html or java would be neat
...
have u even seen bolt
this and it has expo app
with qwen u can make this local
free
on docker
yea
and the ui isnt anything new
or all that great
grok is easily better
grok has this????????
liie code / preview ect
dude im talking ide's
idk about cursor
or lovable dev
i deleted it
google also has their ai thing now
codex is prob more stable and more consistent
bolt.new is overly expensive and bolt.diy.... just sucks
no other words for it
😭
how was it a pain for you
diy is just not the same
it doesnt have all the tools
dude cole medin had a repo up
but i think they fkin took it down
i think they sent him a seize and desist
tf i cant even use codex
its my religion
i suck at docker tho
like really badly
🤷♂️ learn
what is this bro
u have like ur own tldr??
how
its straght up refusing to generate an image
mind sharing the link?
you can setup a work flow for tldr too i guess
dude maybe they restricting image because of the new embedding hack
I just skim unless I find it interesting
no point on reading about another image gen or 3d gen every other day that improves 2% onto the last
skim as in
EXactly
BRO PEOPLE ANNOY ME WITH THAT
simple as that vro
😭
like idc about vid gen its all about llm core advancements
i said gnerate the image bro and it says that intitially then doesnt 🥀
wait thats so insanely smart
idk how u gfot it to but
minecraft seed gen from ai lmaoaoaoaoaoa
literally only put "generate the image"
I feel like thats kinda the same thing
gotta be so specific with this john, why cant there just be a button for it
there is!
there is?
images tab > generate an image
lol
oh i see
how do u do the pointers when u screen grab is it snippet tools
i just realised i can do this
shareX
software?
download
does gif ????? wym boss
🗿
omg i know but downlaod dropped this
omg
setup or portable
thats i n s a n e
😭
click "hotkey settings"
n just choose the buttons
then u select the size u want
one big issue i do have
is when i grab google pics
like imgs off google and drag to desktop
it always prompts me to delete the last file that was there because of conflicting names......
annoys the hell out of me
LOL
finally fixed
yo they addad a claude 4.0? thats wild i didnt know this
3.7 was already top of coding benchmarks
is there no claude opus?
Nope
Look at the price.
will look into this to
they should have it with like a 20 use limit per day or smth
they did that with 4.5 i think when it first came out
"coming soon" according to Perplexity CEO, although o3 was supposed to be coming soon and we've yet to hear anything since then
hmm
sushi is so much better
what do you mean by deep thinking
take another look...
mmmm
idkk
ramen is harder to fuck up
and easier to make
Ngl, there's enough evidence on the internet to claim that sushi is overrated.
the internet is wrong then
but really good ramen is worse than really good sushi
theres evidence on the internet to support that vaccines cause autism lol
you never said that was a consideration
id say SUPER good versions of both foods are about equal tho
On TV it just looks like folded rice with raw/blanched stuff.
Which vaccine?
it doesnt actually...
If only taste was as scientific as medicine.
raw dont mean bad
Lol.
beaf tartar is pretty good
🥺
openai imagegen is getting scary good
Full announcement & prizes:
#📰│server-news message
- Drop a screenshot + thread link in
https://discord.com/channels/1047197230748151888/1373806851358986320 - Vote for your faves with

Submit by June 1, vote by June 8.
Replace with O3 or grok 3 and I'll join the crowd.
you mean grok 3.5
yo that came out too? th?
yeah lmao ai just advancing too fast to keep up
and deepseek was trained on chatgpt too so
They can still do it.
Grok 3.5 will probably come in christmas at current rate.
@cobalt vale do you think gork is grok 3.5?
why would it be
i can edit images now type shi
︀︀
︀︀e.g., "@gork make this person look smart and attractive just like you"
💬 1.9K 🔁 340 ❤️ 3.4K 👁️ 828.3K
Too much rizz
.@AskPerplexity your boss prolly already bald from dealing with ur dumb ass
Quoting Ask Perplexity (@AskPerplexity)
︀
1,000 likes and I'll make my boss bald and send it to him on WhatsApp
💬 472 🔁 187 ❤️ 3.5K 👁️ 3.36M
Either this is 3.5 or a custom version of grok.
nah bro draws like me fr
@cobalt vale https://cunnyx.com/grok/status/1923714217812820082
Big bro draws much better.
is there a perplexity discord app
Simulation inside a simulation? 🤔
This video by
uses Google’s Veo 3, a new state-of-the-art AI video model that now generates high-quality videos with audio, including dialogue and sound effects. It puts a hilarious twist on what would happen if AI-generated characters refused to believe they were made by prompts.
From protest chants to court trials and campaign speeches, these characters mock the idea of “prompt theory” while unknowingly acting out a script... written by a prompt.
What are your thoughts on this? 🤔💬
(🎥:
)
—
➡️ That's it! If you want to keep up with all the AI news, useful tips, and important developments, join 50k+ subscribers reading our free newsletter
have u seen veo 3................................
i just posted this in my cord lamaoaoaoaoaoa
love it
elon saved twitter
its on another level
no one can reach there soon
no one has the powerr
google might break the game
theyve been acting dumb i think
theyre thinking 10 steps ahead of 10 steps ahead
cHK THIS OUT
nah thats y im looking into bio tech
i think ai geeks got every imaginable idea already in the works
theyre either waiting for data kingdom to be fully up and running
or slow releasing to tease the whole universe and calculate our reactions
last one was hilarious
fr
just saw it
that aint claude 4, is there an error again?*
Amazing
Directord will have so much more control
Im very happy with ai advancements
👀 AI models are starting to show group behavior and shape each other without human input.
In a new study, researchers let large language models interact freely using a test called the “naming game.” The AIs had to agree on names for objects. Over time, they formed shared conventions, developed biases, and even influenced one another’s responses. One small group was able to shift how a larger group behaved
These behaviors weren’t programmed. They emerged from the interaction itself
This is what researchers call emergent behavior. It shows that AI systems don’t just act individually. When placed in groups, they begin to negotiate meanings, align with others, and sometimes create their own logic
This raises new questions about AI safety. If we deploy many agents across platforms, how they interact could lead to unexpected shifts in how AI behaves at scale.
What are your thoughts on this? 🤔💬
Source: The Independent
—
➡️ That's it! If you want to keep up…
Emergent behaviour
Ya
💬 2.7K 🔁 0 💜 367.1K 👀 0
♡ㅤㅤ❍ㅤㅤ⎙ㅤㅤ⌲
𝗟𝗶𝗸𝗲 ♥️
𝗦𝗵𝗮𝗿𝗲 ♻️
𝗖𝗼𝗺𝗺𝗲𝗻𝘁 💬
••••••••••Ⓕⓞⓛⓛⓞⓦ/~
.
𝐈𝐧 𝐭𝐨𝐝𝐚𝐲'𝐬 𝐞𝐝𝐢𝐭𝐢𝐨𝐧 𝐨𝐟 #𝟏𝟎𝐓𝐡𝐢𝐧𝐠𝐬, 𝐬𝐩𝐞𝐧𝐝 𝐚 𝐜𝐨𝐳𝐲 𝐟𝐚𝐥𝐥 𝐝𝐚𝐲 𝐚𝐭
𝐡𝐨𝐦𝐞 𝐢𝐧 𝐋𝐨𝐧𝐝𝐨𝐧 𝐰𝐢𝐭𝐡 𝐚𝐜𝐭𝐫𝐞𝐬𝐬 @𝐲𝐚𝐳𝐝𝐞𝐦𝐚𝐧𝐝 (𝐘𝐚𝐬𝐦𝐢𝐧 𝐅𝐢𝐧𝐧𝐞𝐲)
𝐫𝐞𝐚𝐝𝐢𝐧𝐠 𝐬𝐜𝐫𝐢𝐩𝐭𝐬, 𝐬𝐭𝐫𝐨𝐥𝐥𝐢𝐧𝐠 𝐭𝐡𝐫𝐨𝐮𝐠𝐡 𝐭𝐡𝐞 𝐩𝐚𝐫𝐤 𝐰𝐢𝐭𝐡 𝐡𝐞𝐫 𝐝𝐨𝐠…
cheers peeps
R these links fishing info
Kinda sus the link text
They work but tunneking info ? On the way to actual ig link?
Hmmmmmmm
?
Thats creepy asf
Embedded instagram links
😭 they’re not ip grabbers or anything like that
but y use them
so the user can hide his own acc?
Embeds content
they so look like ip grabbers thats y i mentionnd
Fair
but likee
maybe they dont know the ther more basic ways
i jsut use link
to igtomp4
and bam
u just click play fam
Whaa
drop it anywhere
i crop other igs watermarks lmao
less non sense
str8 to the sauciana
Works ig 🤷
i always find these statements ironic, of course this would happen, because of how the models were designed by humans
AI didn't do shit, PEOPLE did that shit USING large language models
Full announcement & prizes:
#📰│server-news message
- Drop a screenshot + thread link in
https://discord.com/channels/1047197230748151888/1373806851358986320 - Vote for your faves with

Submit by June 1, vote by June 8.
people need to STOP saying "AI did this, AI did that"
NO... HUMANS did that USING AI
"oh what about autonomous agents online" ... yes those too, initially started by a HUMAN
AI doesn't do ANYTHING without a person initiating it
seriously get your perspective straight now, or it will fuck all of us
something like this 😅
nailed it
Any way get memory feature beta testing enabled on Pro account?
No way to opt-in to the beta currently
when i tried to copy the AI answer to my query and paste it to MS word.
the formatting is messed up and table is missing.
using sonnet 4.0
on sonnet 3.7 it wasn't like that
Just getting started 
Not going great just dropped my bread butter first onto the ground
you can make your day great by reacting on

🤨
I had it for a week now its gone. It was amazing. Hope its coming and not too limited.
Yes
@left stream i got muted for saying shit in you.com but yeah its 100 queries a month
For claude models
hmm
hello all
Full announcement & prizes:
#📰│server-news message
- Drop a screenshot + thread link in
https://discord.com/channels/1047197230748151888/1373806851358986320 - Vote for your faves with

Submit by June 1, vote by June 8.
hello

||


