Since OpenAI launched ChatGPT in 2022, artificial intelligence (AI) has reshaped the way humans think and operate. Contrary to fears of an AI takeover, humans and this emerging technology are finding ways to coexist productively.
OpenAI’s ChatGPT launch, however, has intensified competition with tech giants, particularly Google, whose AI response, Gemini, has been playing catch-up.
The beginning: ChatGPT’s breakthrough
OpenAI released ChatGPT to the public on 30 November 2022, attracting over one million users within five days and marking the beginning of the generative AI revolution. By January 2024, it had amassed 100 million active users, making it the fastest-growing consumer application at the time.
OpenAI was founded in 2015 by researchers, academics, and entrepreneurs, including Sam Altman, current CEO, Greg Brockman, and Elon Musk. Initially, it secured $1 billion in funding.
Though the company had been developing AI models for years, the release of GPT-3 in June 2020 laid the foundation for the generative AI boom. However, it was ChatGPT’s ability to engage in conversational interactions and answer complex queries that captured global attention.
By March 2023, it announced the launch of GPT-4, which it claimed exhibited human-level performance on various professional and academic benchmarks, from the US bar exam to SAT school exams. Since then, it has launched several upgrades, culminating in the recent image-generation update that allows users to instantly create and modify images.
Previously, users relied on DALL·E, but OpenAI has now integrated advanced image-generation capabilities directly into ChatGPT using GPT-4, elevating the capacity of its AI model.
Also, its partnership with Microsoft has positioned OpenAI as a force in the enterprise AI market, with OpenAI models now integrated into Microsoft cloud services and products, such as Copilot in Office 365.
Read also: OpenAI launches new AI agent to speed up research
Caught off guard: Google’s response
Although Google pioneered the transformer architecture underpinning modern large language models (LLMs), it was blindsided by OpenAI’s ChatGPT launch. Google’s research paper in 2017 introduced the transformer model, yet it had primarily integrated the technology into search rather than developing a standalone AI-powered chatbot.
The transformer architecture helps AI models focus on the most relevant pieces of information they’re analysing. OpenAI’s breakthrough forced Google into urgent action, and according to Bloomberg, Google reassigned more than 1,000 engineers, about 20 percent of its search engineering team, to generative AI projects. A Wired report revealed that Google set an internal deadline of 100 days to develop a ChatGPT rival, leading to the launch of Bard.
Built on Google’s LaMDA (Language Model for Dialogue Applications), Bard underwent rapid development, with 80,000 employees contributing to its creation. However, before Bard’s full release in March 2023, OpenAI had launched GPT-4, which surpassed LaMDA’s analytical and coding capabilities. By the time Bard was available in the U.S. and U.K., it was useful for writing emails and research papers, tasks that ChatGPT had already excelled at.
This didn’t deter Google, which immediately shifted focus to a new language model, Gemini. This model resulted from the merger of the company’s two world-class AI research teams, DeepMind and Google Brain, leading to the formation of Google DeepMind. Gemini had to be the model that would help Google catch up, according to reports.
When Google unveiled Gemini in December 2023, it outperformed ChatGPT in 30 of 32 standard AI benchmarks. Gemini Ultra, the most advanced version, outperformed human experts to achieve a 90 percent score on MMLU (Massive Multitask Language Understanding), which evaluates AI capabilities across 57 subjects, including math, physics, history, law, and medicine.
Gemini also introduced multimodal capabilities, allowing it to analyse research papers, YouTube clips, and other complex data formats. Gemini ended Bard, and Google has since integrated its AI model across its suite of products, including Search, Gmail, and Workspace.
However, Google is still playing catch-up to OpenAI. According to data from Sensor Tower, OpenAI’s estimated 600 million all-time global app installs for ChatGPT dwarf Google’s 140 million for the Gemini app.
Ongoing rivalry
The AI race is far from over, with new capabilities continuously integrated into existing models and their upgrades. ChatGPT’s image integration has been the talk of the moment, with Sam Altman, OpenAI’s CEO, tweeting, “The ChatGPT launch 26 months ago was one of the craziest viral moments I’d ever seen, and we added one million users in five days. We added one million users in the last hour.”
Google also rolled out Gemini 2.5 recently, which it termed its most intelligent AI model, capable of reasoning through its thoughts before responding. This results in enhanced performance and improved accuracy. It outperforms other AI models, including ChatGPT and DeepSeek, in various tests and is currently leading in common coding, math, and science.
However, according to experts, Google’s biggest letdown was allowing ChatGPT to become a household name before staking its claim in a field it should have dominated, considering the company had been touting AI advancements more than six years before ChatGPT’s launch.
Google’s consumer experience has also been called out. While Google’s models have demonstrated impressive technical capabilities, OpenAI’s ChatGPT is widely regarded as more user-friendly. Nikunj Kothari, a product lead at Meter, tweeted, “This is the core problem with Google, they can build the best models in the world, but if they don’t focus on the consumer experience, they are going to get MOGGED again and again.”
While ChatGPT and Gemini are the early leaders in AI innovation, other chatbots including Claude, Copilot, Grok, DeepSeek, Llama, and Perplexity are playing catch-up. There is also the growing concern about the sustainability of generative AI systems, which have required billions of dollars in investment and vast amounts of energy.
The next frontier in AI appears to be agentic AI, which enables systems to perform complex tasks autonomously. With its upcoming Operator service, OpenAI’s model will allow AI to interact with websites, execute tasks like booking travel, and fill out forms. Google is also working to integrate agentic features into its upcoming Gemini iterations.
