Exploring ChatGPT for Improved Observability
Overview
Syllabus
intro
preface
why do we care about observability?
modern platforms are ephemeral
cost of unexpected downtime has risen significantly
todays observability challenges
risk of it outages are set to grow even further in 2023
modern observability solutions
there is currently a lot of innovation occuring in ai, in particular llm's
imitate brain-like functionality using deep neural networks
dnn models are pre-trained from "the whole internet"
chat gpt
chat gpt is built on top of latest breakthroughs in language model design
chat gpt ≠skynet
use cases of observability
keep in mid that chat gpt's responses are non-deterministic
chat gpt's informed decision making
prompt engineering is a new discipline
hyperscalers are investing heavily into gpt's
so are large models a panacea?
dynatrace - cloud done right
Taught by
Conf42