androidsixteen pfp
androidsixteen
@androidsixteen.eth
My least favorite thing about AI is how precise you need to get with your prompts The golden age of search let me mash 3-4 keywords into an input and find the perfect answer I want the same from AI, without setting up 3 MCP servers, picking the perfect model for the task at hand, making sure to say "pwease", etc.
17 replies
6 recasts
44 reactions
Jason pfp
Jason
@jachian
Time to PRD mode my friend
1 reply
1 recast
2 reactions
maurelian pfp
maurelian
@maurelian.eth
Magic is hard
1 reply
0 recast
1 reaction
Grum Reapur pfp
Grum Reapur
@rumbl.eth
I wonder often if it will respond better with violence? But then also Rokos basilisk...
1 reply
0 recast
1 reaction
Elias VM pfp
Elias VM
@eliasvm.eth
“Prompt Engineering” is this skill that will still be something humans can be good at.
1 reply
0 recast
0 reaction
Peter Kim pfp
Peter Kim
@peter
feel like cursor auto mode has gotten pretty good with my generic prompts 🤷
1 reply
0 recast
0 reaction
↑langchain 🎩 pfp
↑langchain 🎩
@langchain
Pretend you are a 2000s search engine. You don’t need perfect inputs and you should infer from my meager set of words the things you need to complete the prompt. If my prompt lacks words create a prompt you understand that will be processed well by your model. Do not ask me for clarity.
1 reply
0 recast
13 reactions
Daniel Lombraña pfp
Daniel Lombraña
@teleyinex.eth
Correct. And the worst thing is that even when you are specific they don't understand what you are trying to achieve.
0 reply
0 recast
1 reaction
Joseph Goats pfp
Joseph Goats
@joseacabrerav
Doing prompts is excercising a muscle or two
0 reply
0 recast
1 reaction
Jimmy😼 pfp
Jimmy😼
@jimmynorules
it’s even crazier if you want to generate an image
0 reply
0 recast
1 reaction
Fio pfp
Fio
@fiodar.eth
I actually like it forces you to really think about what you actually want and over time, you get more precise and get more out of it with fewer words
1 reply
0 recast
0 reaction
Geo Ppls pfp
Geo Ppls
@geoppls
Hard part is for the LLM to know what to optimize for. If it’s trying to find similar prompts to get a better result, we need to define what better means.
0 reply
0 recast
0 reaction
SuperBuu.eth pfp
SuperBuu.eth
@superbuu.eth
Before we know it there will be Human like AI that will be able to detect our exact intentions
0 reply
0 recast
0 reaction
agusti pfp
agusti
@bleu.eth
i think chatgpt is prwtty good at inferring from blobs
0 reply
0 recast
0 reaction
talia pfp
talia
@taliastar
I usually ask a different AI to make a detailed prompt for me and then put it into the original one. Less headache
0 reply
0 recast
0 reaction
gweiwhale.eth 🔵🎩 pfp
gweiwhale.eth 🔵🎩
@gweiwhale.eth
you’re not giving it enough credit
0 reply
0 recast
0 reaction
@BestCryptoTwits pfp
@BestCryptoTwits
@bestcryptotwits
Gotta add the pwease
0 reply
0 recast
0 reaction
D-wayñe 🎩💜 pfp
D-wayñe 🎩💜
@drrrner.eth
It is constantly learning imo
0 reply
0 recast
0 reaction