As we know, ChatGPT was launched by OpenAI in late November 2022. It was a pathbreaking event. Google was already testing generative AI for several months by then. All that led to various models emerging from different divisions within Google. None was good enough to excel GPT-4. Google deferred its plans to launch a rival model, while sorting out the scrambled research work. In the meanwhile, it released a chatbot called Bard. However, it was considered less sophisticated than ChatGPT.
A year later, Gemini was ready but some flaws were detected in it in image generation. The release was delayed, and the company could not seize the opportunity to be a leader in generative AI, the technology which got shaped in the lab of Google. (Ref. Attention Is All that You Need by Vaswani et al in 2017).
Google enjoyed leadership position in internet revolution with its search engine — Google Search. (late 1990s and early 2000s). Later Google diversified into mapping, email and more to become the most valuable company in 2016.
ChatGPT was another event after 25 years of Google’s launch. It was a tool to navigate the online information more creatively. Microsoft was determined to take advantage, while Google stumbled. Microsoft tied-up with OpenAI and funded its research. It embedded its existing products with AI. By doing so, it has become the most valuable company in the world.
After initial hiccups, Google is steadying itself. Gemini has become acceptable in tech circles. Google is thinking of adding paid generative AI services to its Search engines. However, Google so far has earned most of its revenues through advertising. It is struggling to become successful in generative AI space.
There are several flaws in Google’s strategy — there is lack of a corporate plan for rolling out generative AI. The company has fragmented organization structure. There could be simmering inter-departmental tensions. Google has to master execution of its strategies. These are, however, early days. Google is well-positioned to move ahead.
At present, the formidable image Google has in search engine space brings outsized attention to its minor flaws. Google is addressing cultural and organizational issues. Sheer size of Google also causes certain problems.
Google has merged two research divisions — DeepMind and Google Brain. It is now Google DeepMind. There should be coordination between the research teams.
The new technology could cannibalize its traditional search business. Generative AI is direct, and the search links displays require further effort. It has to protect its cash cow product.
Google cannot afford to ignore generative AI. The longer it takes to adopt it fully, the greater is the risk of consumers switching over to rival companies.