#Reproducing the results in the "GPTs are GPTs" paper. Help with ONET labeling

5 messages · Page 1 of 1 (latest)

glad inlet
#

Context:
Hi, I am a Data Scientist in working on a local reproduction of the "GPTs are GPTs" paper (https://arxiv.org/pdf/2303.10130.pdf). In this paper, the authors discuss the impact of LLM on the workforce and give predictions on the expected levels of automation.

I would like to reproduce these estimates on local data, unfortunately the authors do not provide the occupation labels generated by GPT (estimates on the level of automation potential per task for each occupation).

Question:
Therefore i need to generate these myself! That means calling the GPT-4 api with a categorization instruction (rubric) on 20 000 Occupation/task combinations (approximately 10 tasks per occupation or so). Since the rubric described in the paper is about 6000 characters, this operation can be quite expensive. I am therefore wondering if any smart people in this forum could assist me in making smart API calls. From what i understand, i would have to pass the rubric with each request, costing nearly 1500 tokens(!), even though the task/occupation pairs are only about 100 each. I only need a 2 token response for each label (2 characters). Is there a way i can make the request to GPT-4 (Or any other, more suitable model(?)) where i only need to send the task/occupation pairs each time for a label?

Thanks in advance! cursed_thumbsup

wintry falcon
#

This is not time-sensitive GPT work. Pease write a script to run your openAI API calls only during the night or during other non-peak times.

#

As for prompting, just ask ChatGPT to "Make a list table displaying the automation_potential_estimate for each task in the following list of tasks(profession)" it will generate a 2-column list. You can expressly ask to add a number-index column on the left side. Etcetera. You could also ask Chat GPT to generate a number-indexed csv tuple sequence...

glad inlet
#

Wow, thanks! I'm in Europe, so what would equate to night time or non-peak times globally?

wintry falcon
#

There is a website tracking openAI API response time. But, "off peak hours" may be mythological at this point.