What is GPT?
GPT is a general-purpose technology. Machine Learning is being touted as a general-purpose technology. Many other emerging technologies also have received similar hype. However, general-purpose technologies are typically identified with the benefit of many years of hindsight.
Leaders before deciding on any technology strategy involving new technologies (which are generally hyped) and before taking any important decisions, need to assess and understand whether an application or technology is general purpose or not.
There are many research papers on the topic. The research results show that machine learning and related data science technologies are relatively likely to be general-purpose. There is a strong overlap between six technologies: ML, business intelligence, big data, data mining, data science, and natural language processing (NLP). Of these six technologies, ML is the technology most hypothesized to be a GPT. However, the research results suggest that the other five technologies also demonstrate and display similar patterns and together are relatively likely to represent a GPT.
Gaining competitive advantage from GPTs requires investment in internal innovation capabilities and collaboration with academia and other organizations at the technology frontier, as well as an ability to sustain financial investments with long-return horizons. However, resolving this speculation using existing empirical methods for identifying GPTs requires many years of hindsight.
By that point, the managerial relevance of knowing whether ML, blockchain, cloud computing, and other technologies are GPTs is limited. GPTs are transformative because they open up new opportunities for innovation and economic growth, linking the technical implementation of innovation to its macroeconomic consequences.
Accruing GPT benefits requires complementary innovations that take advantage of the capabilities of the technology (e.g. Greenwood and Yorukoglu 1997; Aral, Brynjolfsson, and Wu 2012; Tambe, Hitt, and Brynjolfsson 2012).
As a consequence, the benefits of a GPT are large but occur with a long lag (Bresnahan, Brynjolfsson, and Hitt 2002). They also typically require long-run financial investments. For example, electricity’s early uses focused on street lighting and street railways (Lipsey, Carlaw, and Bekar 2005). Over time, innovation occurred in a wide range of sectors, from upstream advances in power generation to downstream development of household appliances such as washing machines, vacuum cleaners, and refrigerators. Electrification also led to the re-organization of factories (David 1990).
Importantly, it was primarily these later innovations that drove productivity growth both within companies and at a macroeconomic level. Therefore, it is useful to have an early sense of whether an emerging technology is likely a GPT for leaders to make informed decisions about their organizations' technology strategy.
(Above source: Could machine learning be a general-purpose technology? SSRN Electronic Journal 2020 - Avi Goldfarb, Bledi Taska and Florenta Teodoridis).
What is GPT in ChatGPT?
Generative Pre-trained Transformers (GPTs) are large language models (LLMs). There are large language models and there is LLM-powered systems.
Some of the key findings of the paper published on "An Early Look at the Labor Market Impact Potential of Large Language Models" by Tyna Eloundou1, Sam Manning1,2, Pamela Mishkin∗1, and Daniel Rock3 are as below:
That around 80% of the U.S. workforce could have at least 10% of their work tasks affected by the introduction of LLMs, while approximately 19% of workers may see at least 50% of their tasks impacted.
They do not make predictions about the development or adoption timeline of such LLMs. However, the projected effects span all wage levels, with higher-income jobs potentially facing greater exposure to LLM capabilities and LLM-powered software.
Significantly, these impacts are not restricted to industries with higher recent productivity growth.
With access to an LLM, about 15% of all worker tasks in the US could be completed significantly faster at the same level of quality.
When incorporating software and tooling built on top of LLMs, this share increases to between 47 and 56% of all tasks.
LLM-powered software will have a substantial effect on scaling the economic impacts of the underlying models.
Conclude that LLMs such as GPTs exhibit traits of general-purpose technologies, indicating that they could have considerable economic, social, and policy implications.
They go on to add that when regressing exposure measures on skillsets using O*NET’s skill rubric, they discover that roles heavily reliant on science and critical thinking skills show a negative correlation with exposure, while programming and writing skills are positively associated with LLM exposure.
While the requirement for human supervision may initially slow down the speed at which these systems diffuse through the economy, users of LLMs and LLM-powered systems are likely to become increasingly acquainted with the technology over time, particularly in terms of understanding when and how to trust its outputs.
More details about the report can be obtained from the following link: https://arxiv.org/pdf/2303.10130.pdf
To conclude GPT4 is a potential disruptor to the labour market, not just in the US but in the entire world. Its already attained the tag of a general-purpose technology. The capabilities include text input via ChatGPT and the API. To prepare the image input capability for wider availability, they are already collaborating closely with a single partner to start.
The above article is published for informative purposes only.
Comments