With signs that there’s plenty of cash to go around for generative AI startups, Gartner placed genAI on the top of the ‘Peak of Inflated Expectations’ for the first time. If you think this technology is overhyped right now, wait for another 2-5 years, as per the report. 

Startlingly, ChatGPT, the most well known example of genAI, already crossed the threshold of a product life cycle within the first three months of its release as per previous reports. The rapid adoption rate can be attributed to the media spotlight that further amplified the service expectations. 

Apart from being flooded with investments, the technology is also being looked up as a driving force to revamp the industrial landscape by tech giants. Leading software corporation IBM is one of them as it paused hiring for 7,800 jobs due to AI. The company’s CEO Arvind Krishna believes AI’s ability to increase worker productivity is the solution to the talent crunch problem faced by many. Like others, the company also jumped headfirst into genAI with its in-house Watson assistant. 

Another factor that will help the technology reach its peak is the sudden outburst of open source models. From Meta’s Llama 2 to Databricks’ Dolly, iterations of open and free language models have been making headlines daily in the recent past. Alongside, Hugging Face, the messiah of open source, is hosting an extensive number of diffusion as well as language models. With quality resources from every corner, contributions from the FOSS community is laying the foundation to the upcoming peak of the genAI. 

ChatGPT is not everything 

Generative AI is not a singular technology, stated Gartner analyst Arun Chandrasekaran in an interview. It includes everything from foundation and diffusion models to prompt engineering tools. “They all enable this trend of generative AI,” he elaborated. 

Released two days ago, a report highlighted that every piece of visual art humanity has created over the last century and a half has been outnumbered by AI generated art just within a brief span of 1.5 years. The major chunk was produced by open-source diffusion models — developed last year by Stability AI. 

For a transformer based technology like GPT-n, the ‘n’ number of identified use cases and billions of dollars channelled is the reason for its fame. While previous generations of the software could technically do these things, the quality of the outputs was much lower than that produced by an average human. A rarely mentioned advantage of the AI is that it’s remarkably good at things that would take humans decades to do, like processing the entire canon of a certain literary or learning an artistic movement. 

According to Chandrasekaran, the main reason AI has hit the peak of the hype cycle is the sheer number of products claiming to have generative AI baked into them. Startups started flocking to the generative AI boat after OpenAI introduced GPT4’s capability via ChatGPT and the API. Even Sam Altman, the CEO of OpenAI has stated in the past of the technology becoming ‘wildly overhyped’. 

Technology’s ever-evolving landscape is a tale as old as time. In the early 1990s, there was the advent of the Internet and the opportunities it presented for enterprising startups to revolutionise industries ranging from e-commerce to software downloading. 

As the landscape shifts once again with generative AI, Startups but veterans like  are scrambling to find ways to leverage these cutting-edge technologies to gain a competitive edge. Even veterans like Google and Microsoft have redoubled their efforts with red teams to make the most of the technology. 

Embryonic stage 

While some have declared the technology as an inflating bubble, on the contrary as per Gartner it is still in the nascent stage. Aligned with the Gartner study, analysts predict the return on investment on the technology after 2 years. The hard part for the quality content producing technology is turning the value of time and productivity into an ROI measurement. 

Further proof of concept metrics as per experts also include scalability, ease of use, quality of response, accuracy of response and explainability or total cost of ownership. 

McKinsey estimates that the technology could add between $2.6 to $4.4 trillion of economic value annually across various industries. The study titled ‘The economic potential of generative AI: The next productivity frontier’ draws results from 63 new use cases analysed across 16 business functions that could deliver those returns. These estimated returns are comparable to the UK’s 2021 GDP of $3.1 trillion.

While the ROI egg remains unhatched, companies should be focusing on tailoring and putting the technology to use as per their respective products. Trying to jump on the bandwagon for the sake of it can result in an actual bubble resulting in a market crash if not thought through.