A look into ChatGPT and its possible future applications.
Late 2022, OpenAI released ChatGPT and it has been one of the most talked about topics since, including in the oil and gas industry. ChatGPT utilizes language processing to create human-like conversational dialogue. It derives language models that respond to clearly defined questions and generates articles, social media posts, essays, emails, and even programming codes. The prerequisite of large training datasets lends itself to utilization in the oil and gas industry. For example, early June, Vital Energy announced that they expect to see a 2-3% increase in total field production, 15% increase in gas-lift runtime, and a 4% increase in ESP runtime by implementing AI technologies.
A company with large datasets like TGS is in a unique position to benefit from AI models like ChatGPT. Machine learning has already been successfully applied; currently, TGS predicts missing log data mainly on-shore US in TGS’ ARLAS (Analytics Ready LAS) product. ARLAS uses a gradient-boosted regression tree to generate a full quad combo log suite incl. P-sonic for every well in 17 major unconventional basins in the US. The logs provide valuable information to interpreters and technicians and are being used to map structure and predict reservoir properties in areas where the actual measurements are sparse.
Applications for ChatGPT-type models are easy to imagine. They will allow us to make quicker decisions regarding which acreage to lease, where to drill, how to drill, how to produce, how to distribute the product, etc. The key is the quality and quantity of the data that are being used by AI; more data and better quality are essential. In addition, new cybersecurity regimes must be developed and applied to make this a secure tool that protects individual company IP. TGS uses data architecture such as MDIO to facilitate effective, cost-efficient, and secure development of AI models to produce insights that are not available through conventional products.
For more information on TGS Products, to discuss how TGS uses AI, or to schedule a demo, contact us at WDPSales@tgs.com.
Image 1: Permian Basin LAS: measured sonic logs
Image 2: Permian Basin LAS: ARLAS predicted sonic logs using AI
