One year of ChatGPT: Is an AI winter in the offing?

Last week, the AI ecosystem of the world was rocked by unprecedented events as the most prominent person in the AI universe, Sam Altman, was sacked as CEO from OpenAI.

One year ago, Altman made history by releasing the AI-based chatbot ChatGPT. His product caused an all-out disruption in the world of technology with tech giants such as Alphabet Inc, the parent company of Google, rushed to update their product to compete with OpenAI’s chatbot.

Subsequently, Google released its own AI chatbot called Bard. Moreover, other tech giants such also added AI-enabled services given the threat posed by OpenAI’s revolutionary product.

ChatGPT: the promises, pitfalls and panic

Fast forward one year, Sam Altman is one of the most prominent personalities of 2023. Coming back to AI, first world countries realized the threats posed by AI to labour, humans, security, privacy and jobs and introduced AI regulations to limit its functions. Once again Sam Altman played a solid role, being at the centre of AI growth.

Last week, he was fired by OpenAI’s board citing inconsistent communication and inability to handle responsibilities. While the board of OpenAI hired Twitch Co-founder Emmett Shear as the new CEO and Altman was hired by Microsoft, OpenAI’s reputation took a drastic hit and it attracted massive criticism for sacking the person who led an AI revolution.

After majority of the staff threatened to leave OpenAI, Altman was reinstated as the CEO.

This pointed to the extent of power held by Altman in the world of technology and AI. He has the potential to become one of the most powerful tech CEOs akin to Mark Zuckerberg, Elon Musk, Satya Nadella, Tim Cook and Jeff Bezos.

The jitters seen in the world of technology over the past few days reflects the important of just one person in the AI world.

Tech rivals chase ChatGPT as AI race ramps up

Sluggish progress

On one hand, we saw the OpenAI-Sam Altman fiasco and on the other, the growth of AI products has come to a standstill all over the world as there have been no new visible advancements.

The last huge thing in AI was ChatGPT. Although OpenAI released GPT 4 as an update to GPT 3.5, it is yet to start work on GPT 5.

Moreover, the leading two generative AI products, ChatGPT and Dall-E, are both made by OpenAI. In summary, other companies are falling far behind in development of AI tools and Open AI lacks resources, such as servers, to expand its AI portfolio.

In fact, OpenAI is currently reporting a loss due to high of running the servers that are training ChatGPT. Following ChatGPT’s release, other tech giants raced to incorporate AI advancement but so far they have only been able to replicate the chatbot and incorporate on their websites.

One underlying problem within AI is that there haven’t been many revolutionary advancements since the release of ChatGPT and there seem to be none planned either.

Thus, there seems to be an AI winter in the making.

Pakistan and the genie of Artificial Intelligence

Problems

One reason behind massive slowdown in AI products is the cost to create and keep them running. To train AI models, companies need massive amount of costly servers and graphic processing units (GPUs).

Moreover, they also need trained manpower which is scarce right now. In addition, a large amount of data is also required which is usually not available.

Take the example of COVID-19 where all the data available is just 4 years old – (from November 2019).

The problem here is that it is difficult to make accurate predictions for a longer terms, for example the next 10 years, with data that is available for just a few years.

Many times, companies are forced to be patient to train a model. Similarly, most AI pre-requisites and resources come with lofty costs that encourage companies to shelve their plans.

In essence, an AI winter is in the making but it is hardly the fault of the industry.

The article does not necessarily reflect the opinion of Business Recorder or its owners