"human in the loop" markdown "human in the loop" "let's ask Claude" LLM generates docs for the LLM to consume deepseek is pretty good now "llms are just linear algebra" ralph loops "check" output of LLM using another LLM adversarial process prompt injection agent deletes the tests its "agentic enginneering" not "vibe coding pricing comparison between models agent deletes the tests markdown "llms are just linear algebra" adversarial process pricing comparison between models json its "agentic enginneering" not "vibe coding llm also generates the tests prompt injection 𝕏, the everything app llm also generates the tests "check" output of LLM using another LLM "I acknowledge the ethical issues BUT..." "human in the loop" markdown "human in the loop" "let's ask Claude" LLM generates docs for the LLM to consume deepseek is pretty good now "llms are just linear algebra" ralph loops "check" output of LLM using another LLM adversarial process prompt injection agent deletes the tests its "agentic enginneering" not "vibe coding pricing comparison between models agent deletes the tests markdown "llms are just linear algebra" adversarial process pricing comparison between models json its "agentic enginneering" not "vibe coding llm also generates the tests prompt injection 𝕏, the everything app llm also generates the tests "check" output of LLM using another LLM "I acknowledge the ethical issues BUT..."
(Print)
"human in the loop"
markdown
"human in the loop"
"let's ask Claude"
LLM generates docs for the LLM to consume
deepseek is pretty good now
"llms are just linear algebra"
ralph loops
"check" output of LLM using another LLM
adversarial process
prompt injection
agent deletes the tests
its "agentic enginneering" not "vibe coding
pricing comparison between models
agent deletes the tests
markdown
"llms are just linear algebra"
adversarial process
pricing comparison between models
json
its "agentic enginneering" not "vibe coding
llm also generates the tests
prompt injection
𝕏, the everything app
llm also generates the tests
"check" output of LLM using another LLM
"I acknowledge the ethical issues BUT..."
