Enhancing Energy Economics with AI:
Enhancing Energy Economics with AI:
A Practical LLM Workshop for KAPSARC
A Practical LLM Workshop for KAPSARC
Embedded Files
Short Course Description:
This interactive workshop begins with a quick, big-picture introduction to the basics of large language models (LLMs). and then dives into the exciting practical applications of LLM for Energy Economics. Participants will harrness the power of GPT-4 to explore advanced data analysis, analyze financial statements of Energy companies, and efficiently replicate research papers. The course also covers brainstorming, academic writing, and creating custom GPTs, ensuring a mix of productivity and creativity. Above all, this workshop aims to be both informative and fun!
About the Instructor:
I extensively use ChatGPT and other large language models (LLMs) for both research and policy-oriented tasks. I have successfully trained over 300 economists at the European Central Bank, Slovak National Bank, Czech National Bank and KAPSARC. At KAPSARC I am leading a team focusing on implementing AI-driven solutions across the K&A.
Classes: Tuesday November 12
Room: Apex
Seminars: in the afternoon
Teachers: Aleš Maršál
Office Hours:
Ales: at the coffee machine
Grading: Certificate can be issued by me for basket of dates :-)
Requirements
To be able to sign up you need to fill in the survey which helps us to understand if and how employees use generative AI at work
Active participation
Learning Materials
Learning Materials
Other sources:
Generative AI for Economic Research: Use Cases and Implications for Economists by Anton Korinek
Jay Alammar (Saudi super star - illustrative explanation of how LLM works)
Other sources not for free:
People to follow:
What is AI?
The two theories presented in these two articles are two very different concepts how one can look at AI. Please read the article which belongs to your name and fill the 2 questions survey.
read this 2 pager and "Can we reach..." fill this survey
read this two pager and "On the way..." fill this survey
1.1 What is Chat GPT?
By far the best high level introduction to how Chat GPT works is the article provided by Stephen Wolfram. If you want to delve deeper DeepLearningAI courses are good way to go. I can also recommend Jay Alammar book. More technical but nice materials by Maxime Labonne.
2. Literature Survey
Here are the handouts summarizing some of the key papers in Economics and AI.
3. Working with the academic literature
4. Sentiment Analysis
4.1 Fundamental Financial Analysis
Aditonal materials
Yield curve data to demonstrate econometrics using Python programmed by ChatGPT
case study on dropbox, on google drive
Productivity Gains from ChatGPT Use in the Workplace: Evidence from the National Bank of Slovakia :
Productivity Gains from ChatGPT Use in the Workplace: Evidence from the National Bank of Slovakia :
These density plots are results from my research conducted at NBS and illustrate the missive improvement in productivity after incorporating ChatGPT into work.
Density plot showing the productivity improvements of participants in the previous workshop.
This plot compares the overall task scores of participants using AI with those not using AI. The distribution of scores shows that ChatGPT raises productivity levels, especially among lower-performing employees, as indicated by the narrowing distribution in the AI group. This suggests that AI contributes to closing performance gaps while elevating productivity across the board.
This plot demonstrates the impact of ChatGPT on task efficiency, measured in time spent on specific tasks. The blue curve represents the time distribution for tasks completed with AI assistance, while the orange curve shows tasks completed without AI. The shift to shorter times in the AI group indicates that ChatGPT accelerates task completion across participants. The broader spread of the AI curve suggests that higher-performing employees benefit disproportionately from AI assistance, resulting in a widened distribution.
Google Sites
Report abuse