7 Comments
User's avatar
Xian's avatar

“When you read biographies of people who’ve done great work, it’s remarkable how much luck is involved. So you need to make yourself a big target for luck, and the way to do that is to be curious.”

- Paul Graham

ToxSec's avatar

“I’ve seen this happen enough times to know: the model isn’t the variable. The prompt is”

this is absolutely so true. it really does make a difference. it feels like with today’s models if there is a problem it’s either the prompt or context.

great read!

ToxSec's avatar

of course! my pleasure to read these articles Ruben, thank you!

Dhruv Jain's avatar

The move from prompt to context is real. But the next unlock after context is evals. Most people optimize prompts until they get lucky once instead of building the test harness that makes luck repeatable.

Dhruv Jain's avatar

Disagree that prompting is "the most valuable skill." I've built AI workflows for clients and the biggest gap is never the prompt. It's that people don't feed the model real context about their business. One solid system prompt with actual customer data beats 30 clever techniques every time.

Luisma Valencia's avatar

Gracias por compartir todos estos contenidos Rubén. La verdad que la sensación que uno tiene es abrumador ante la cantidad de información y referencias que nos llegan en el día a día. No es fácil a veces abstraerse de todo ello.