The recent discussions about the “AI bubble,” frequently discussed in the tech world, have been exacerbated by the massive investments received by companies like OpenAI and Anthropic. However, Clem Delangue, CEO of machine learning resource center Hugging Face, argues that these concerns are misplaced. Delangue states that the current situation is not a general AI bubble, but rather one encompassing only Large Language Models (LLMs). Speaking at an event, the CEO predicted that the LLM bubble could burst next year, emphasizing that these models are only a small subset of the vast potential of AI in fields such as biology, chemistry, image, audio, and video.
New prediction in the tech world: Specialized solutions are emerging instead of general models
Analyzing the current state of the industry reveals that all attention and financial resources are concentrated in a single area. Investors and tech giants are fixated on the idea that a single, massive model can solve all problems. General-purpose chatbots, often expected to meet everyone’s every need, are at the core of this strategy. However, Delangue is quite cautious about this approach. He disagrees with the idea that a single model built with massive data centers and massive processing power will be a universal solution for all companies and people. Sources like Ars Technica also voice concerns about AI investments, noting that these concerns are generally limited to companies whose core product is LLM.
The future of technology may lie in diversity rather than a single supermodel. Hugging Face’s CEO believes the outcome will be the coexistence of numerous models that are more specialized and solve different problems. Of course, Hugging Face’s role as a library for precisely these specialized models supports this view. The company incorporates large, open-source models from giants like OpenAI and Meta, as well as smaller models that developers have adapted to specific needs. This view isn’t unique to Delangue; research firm Gartner also noted in an April analysis that diversity in the business world is increasing the demand for specialized models and domain data.
Whatever the fate of LLM-based applications, investments in other areas of AI are still in the early stages. For example, it was revealed that former Amazon CEO Jeff Bezos has partnered with a new venture focused on machine learning applications in engineering and manufacturing. This venture, funded by over $6 billion, is one of the most concrete proofs that the term “Artificial Intelligence” is not limited to chatbots. While Delangue’s statements are intended to highlight his own company, they also remind us of an important fact: the concept of artificial intelligence encompasses a very broad spectrum, and we are still too early to see where these technologies will take us.
These rapid changes in the sector and the differing opinions of experts are raising questions about the future of technology. What are your thoughts on this? Do you think the current era is just a bubble encompassing chatbots, or will AI be here to stay, with specialized solutions that will transform every aspect of our lives?
