The article also explores the use of automatically generated prompts in image generation algorithms, with a team from Intel Labs developing a tool called NeuroPrompts that enhances simple prompts to produce better images. Despite the advancements in automated prompt generation, the article suggests that the role of prompt engineers will continue to evolve but not disappear. The complexity of adapting AI for industry needs will still require human involvement, leading to the emergence of new roles such as Large Language Model Operations (LLMOps) engineers.
Key takeaways:
- Since the launch of ChatGPT in 2022, prompt engineering has become a popular method to optimize the performance of large language models (LLMs), with many companies using it to build product co-pilots, automate work, and create personal assistants.
- Recent research suggests that prompt engineering is more effective when done by the model itself rather than a human engineer, casting doubt on the future of human-led prompt engineering jobs.
- Automatically generated prompts, or 'autotuned prompts', have been found to outperform those created through trial-and-error, and have been successfully applied to both language and image generation models.
- Despite the rise of autotuning, experts believe that prompt engineering jobs will continue to exist in some form, as the adaptation of generative AI for industry needs is a complex process that will require human involvement for the foreseeable future.