ChatGPT's one-year milestone: Can Google and Amazon rival its success?
December 8, 2023
Since its introduction, ChatGPT has revolutionized the world of generative AI, but the question remains: Can established tech giants like Google and Amazon withstand the dominance of OpenAI and Microsoft with their latest innovations? Currently, we are observing how the landscape is changing.
Already over a year old, ChatGPT maintains its position at the forefront of the generative AI landscape. Microsoft relies on the underlying language model GPT-4, which is used in its Copilots, including the publicly and freely accessible Bing Chat – now also known as Copilot. Google's efforts to offer a comparable alternative have so far been less successful - to put it mildly. Amazon, on the other hand, has operated more in the shadows.
Now, however, things are starting to change. Amazon Q, a new chatbot specifically developed for business needs, enters the stage. It supports employees in a variety of tasks – from strategy development to software engineering – and enables secure integration of internal data sources. Meanwhile, Google introduced its latest product, Gemini, which will be available in three different versions. The most powerful version, Gemini Ultra, shows promising results in benchmarks and partially surpasses GPT-4. Yet, the question arises whether "slightly better" is enough to challenge OpenAI's dominance, especially since Gemini Ultra is not even on the market yet. After all, the supremacy of ChatGPT & Co. is not based solely on benchmarks.
A crucial advantage of OpenAI and Microsoft lies in their established market presence. Company-internal solutions like myGPT @ Merck, dmGPT, or KaiChat by KPMG, to name just a few, already use OpenAI's language models – switching to other providers is not without effort. Easier access to information about GPT-4 compared to competitors is another factor of success. It's not just about manufacturers providing good documentation. The multitude of best practices that emerge in an environment where many companies work on similar solutions also facilitates development work. It has been shown time and again that providers who have won over developers also win the race for dominance.
OpenAI's developer conference made it clear that the company is planning and already implementing its own ecosystem. The new GPTs and the announced store make OpenAI a platform provider – a model that has proven successful in the past. In contrast, Google had to admit after the presentation of Gemini that certain aspects were staged, and Amazon faces reports of hallucinations and data leaks.
Waiting for real competition
The long waiting times between the announcement of a product and its availability are not unique to Google. Microsoft and OpenAI are also affected. Months passed between the announcement of the Copilot for Office and its availability. Availability was limited to those customers who purchased at least 300 licenses in addition to their enterprise contracts. OpenAI announced ChatGPT Enterprise in August, but apart from a confirmation email, most European companies that have expressed interest have not yet heard from OpenAI.
This situation is crying out for competition. Available, reliable offerings from Google and Amazon could create new dynamics and, in the best case, motivate OpenAI and Microsoft to new heights of performance. As an excuse for not dealing with generative AI and its application in companies, the situation is not suitable. It only shows us that with a watchful eye on market developments, we can reassure ourselves that we are still betting on the right technologies to benefit from the productivity and quality improvements that generative AI already offers today.