The Basics of Prompt Engineering with Azure OpenAI Service

Revolutionize your Enterprise Data with ChatGPT! Did you know the hottest new programming language is English? On this episode of the AI Show, we are excited to welcome Marco Casalaina for a demo on how you can program your own ChatGPT - with natural language. And how prompt engineering really works.
Chapters:
00:00 Welcome to the AI Show
00:26 Welcome Marco Casalaina
00:49 What is ChatGPT and how does it work?
02:04 Bing AI demo begins
08:53 How it works - ChatGPT on top of Cognitive Search
17:48 Wrap
Resources:
Revolutionize your Enterprise Data with ChatGPT Blog
aka.ms/ChatGPT/Blog

[@BijouBakson]

Wow, this is insane! We are at the edge of the computer revolution. The tools we have access to are just mindblowing. The possibilities are endless. Thank you so much for this.

[@cedricvillani8502]

Other modal’s are receiving the output from the original modal as input and there output feeds back into the main model, I’ve created these multi-mini refinement micro services. That’s the shot and sweet explanation

[@dipteshbose]

Oops..that was quite a bit of an uncovering for a new tech like this….thank you so much…

[@verolthompson2779]

Marco, Marco, i'm totally blown away. Does the business/restrictive rules stored the documents ( in this example, the info about your health plan)in table forms before passing the info to your ChatGPT?

[@marcocasalaina350]

It can be stored that way, yes - GPT is remarkably flexible at handling unstructured data in many forms. You'd want to test something like that fairly thoroughly though before releasing it.

[@verolthompson502]

@@marcocasalaina350 Is this an area I can access to play with this tool? If so, how can I? Thanks

[@rakeshyerramsetti3531]

can some one post the article mentioned in the video.

[@solitone]

when the general public will be able to experiment with Azure OpenAI Service?

[@techie1033]

hank you for your application for Azure OpenAI Service. If you submitted an experimentation or production review for Azure OpenAI’s text or code models, we will review your application within approximately 10 business days. ..Seriously ?? I have to wait 10 days.

[@siddhantdhandhania7679]

how can we get enterprise or org id to use this ? anyone ?

[@punisherash388]

I didn't see it coming, lawyers will be best programmera

[@sathyamaniify]

English literatures also.

[@yannolaf]

What may have caused the poor audio quality in this video, and how can it be improved?

[@MONTOD]

I don't have anymore understanding of what is happening by the end of the video 😅

[@CassandraDarling]

This is a badly titled video. It seemed "basic" until suddenly we jump into "I made this bot of ChatGPT on top of Azure Cognitive Search which can use documents you upload to it...." What a sudden and unhelpful leap from novice to intermediate. HOW did you make that bot? How do you layer GPT into your own company's systems? HOW do you upload documents for your bot to use as source material....?" I thoroughly read TWICE the article mentioned. Basically all it says about how to inject your own data is "inject your own data!" FUCKING HOWWWWW. I'm leaving with more questions, without ever really getting the answers to the question I came here for. Just retitle this video more appropriately please.

[@bluesque9687]

We dont want to miss, you say?? Why are there so few views...? Nah..

[@cedricvillani8502]

Lol prompt engineering is not a actual job or ever going to be a marketable skill set

[@digitaltutorials1]

Wrong. There are high paying jobs already.

[@JJ-cf6zn]

Who asked 😂

[@NK-ju6ns]

@@digitaltutorials1 give me one job link for prompt engineer

[@kasforai]

he must be an angry engineer 😂

[@danielarmstrong4941]

Yeah, just like people thought cloud services wouldnt replace on-prem solutions... As businesses implement Azure OpenAI into their own services and processes, they'll want to implement as fast and as concise as possible. Whether you like it or not, this will absolutely become a marketable skill.