OpenAI has just launched an official course —online and (temporarily) free— on artificial intelligence. And it has done so by joining DeepLearning.
AI, the company of Andrew Ng, also co-founder of the Coursera course platform and former CEO of Google Brain (recently merged with DeepMind ) and Baidu’s AI department.
It seems like a good profile to promote a course like this —even if you don’t know that his AI courses launched so far are an international benchmark—, so it’s good news that he himself teaches it together with Isabella Fulford, from the technical team at OpenAI.
The title of this course ( “ChatGPT Prompt Engineering for Developers” ) refers to the OpenAI flagship product and stands for ‘ChatGPT prompt engineering for developers ‘.
What does this course offer?
And, as the name suggests, this 1.5-hour course aims to equip software developers with the necessary skills to integrate large language models (LLMs, such as GPT-3.5 and GPT-4) in the creation of applications.
According to the website, the course is ‘Free for now’, so don’t put it off too long…
Thus, it comes to cover a neglected niche in the market, because despite the rise of this new profession of ‘prompt engineer’, there is a certain lack of quality resources on its application to the field of programming. Now, users will be able to have information not only about how LLMs work, but also about the use cases of the LLMs’ APIs, applied to areas such as
- Summary and transformation of texts (such as the condensation of reviews, translation and proofreading of documents…).
- Information inference (extraction of relevant topics…).
- Expansion of ideas (automatic writing of e-mails…).
In addition to that, the course participants will receive information on key principles to create effective ‘prompts’ for the desired result, as well as to create custom chatbots from a certain LLM (remember that ChatGPT is the chatbot of the LLM GPT-x). To ensure hands-on learning beyond mere theory, the course will also provide a Jupyter Notebook environment where users can explore and manipulate code.
This offers a clue to the only formative prerequisite for this course: a basic understanding of the Python language. This does not mean that it is not very useful for all those engineers with previous experience in machine learning who want to train in the ins and outs of language models.
In the words of Andrew Ng:
“Generative AI offers many opportunities for AI engineers to develop, in minutes or hours, powerful applications that previously would have taken days or weeks. I am excited to share these ‘best practices’ to enable many more people to take advantage of this revolutionary functionality.” “.