Unpacking the AI chatbot race: How have tech giants fared against Microsoft’s ChatGPT?
Almost three months after its creation, Microsoft’s ChatGPT has become the face of technology’s latest innovations in artificial intelligence, launching thousands of high-performance, AI-powered search engines. Given its present-day hype, rival tech companies are working to catch up in the current AI chatbot race, sparking an unforeseen rise in computing power, competition, and technological investment.
ChatGPT is a chatbot launched in November 2022 and developed by an AI research company, OpenAI. GPT, in this case, refers to Generative Pre-trained Transformer — a form of technology pre-trained on large volumes of generic data from texts and images to learn language structures and conversational characteristics. Through Microsoft’s multibillion-dollar investment and partnership, ChatGPT’s technology has been incorporated into their products: Bing and Edge.
Yusuf Mehdi, Microsoft Corporate Vice President of Search, speaking about the integration of OpenAI into Bing and Edge. Photo: AP Photo/Stephen Brashear
This move for Microsoft has redefined the nature of online searches, allowing users to receive highly detailed information on demand, instead of just a list of recommended websites. As Bing further develops into an overpowered tool for research, rival companies like Google have felt under pressure to participate in the race to create the best AI-operated search engine. As Microsoft boss Satya Nadella said, “The race starts today.”
On Feb. 6, 2023, the CEO of Google and Alphabet, Sundar Pichai, released a statement introducing Bard — an “experimental conversational AI service” mimicking the same capabilities of ChatGPT to provide useful, high-quality responses to search queries. However, Bard’s perceived reliability quickly declined as soon as a query about the James Webb space telescope during their demo was found to be inaccurate, leading to losses amounting to $160bn for Google.
Such events have spurred speculation over the accuracy of large language models and the extent to which they could be detrimental to our workspaces. It became clear that rushing AI-powered search engines into the market without extensive specialized testing is one drawback to the fast-paced AI race impacting innovation. Despite Google’s long-standing dominance as the most utilized search engine, Mark Riedl, a professor at the Georgia Institute of Technology, stated, "Microsoft has pulled off a major feat by virtually overnight making search technology a two-way race again.”
Contrarily, Google and Microsoft are not the only companies developing AI-powered products. From a global perspective, the US and China have been long-time competitors in the AI race, hence calling into question the response of China to the West’s recent innovation of language models.
For many years, China has led the race against the US through their numerous patent filings related to artificial intelligence. However, due to China’s strong censorship policies, AI innovation has been brought to a standstill. Chinese authorities have looked down on ChatGPT because of its public accessibility to uncontrollable data. Seeing that ChatGPT-style search engines need large volumes of data to operate, a censored online environment has stifled China from participating fully in the AI chatbot race.
To curb these limitations, Chinese tech giants like Baidu, NetEase, Ali Baba, and JD.Com are working on integrating language models into application-specific scenarios as opposed to an all-around “question and answer” platform like ChatGPT.
Baidu Headquarters in Beijing. Photo: REUTERS/Florence Lo
On Feb. 7, 2023, Baidu announced that they would complete testing of a new product, “Ernie Bot”, in March. Meanwhile, NetEase and its education subsidiary, Youdao, have also been working on generative AI for their educational products. On the other hand, Ali Baba aims to deploy the same technology in their cloud computing products, while JD.Com plans to launch ChatJD — an industrial version of ChatGPT that serves the retail and finance industry. These cases demonstrate how contingent technological innovation can be on a country’s political environment.
Overall, ChatGPT has created a widespread buzz over emerging AI technologies, kickstarting a new era of rivalry among tech companies worldwide. Chatbots are becoming increasingly pervasive, from helping companies like Buzzfeed generate content for quizzes to assisting students with their assignments. Nevertheless, these language models rely heavily on human feedback to improve their accuracy — the more it is used, the more it improves. Therefore, it is important to remain cognizant of critically consuming information produced by these ever-evolving chatbots before they can be fully trusted.