{"id":20698,"date":"2025-01-28T13:00:00","date_gmt":"2025-01-28T11:00:00","guid":{"rendered":"https:\/\/forklog.com\/en\/from-bard-to-gemini-two-years-on-google-still-cannot-catch-chatgpt\/"},"modified":"2025-01-28T13:00:00","modified_gmt":"2025-01-28T11:00:00","slug":"from-bard-to-gemini-two-years-on-google-still-cannot-catch-chatgpt","status":"publish","type":"post","link":"https:\/\/forklog.com\/en\/from-bard-to-gemini-two-years-on-google-still-cannot-catch-chatgpt\/","title":{"rendered":"From Bard to Gemini: two years on, Google still cannot catch ChatGPT"},"content":{"rendered":"

Two years ago Google unveiled Bard, later rebranded as Gemini, as its answer to ChatGPT. The search giant had the makings of a worthy rival: access to up-to-date data, hefty computing power, in-house AI chips, hundreds of specialists and billions in investment. Yet the smaller OpenAI, albeit backed by Microsoft, has kept the lead, leaving Google unexpectedly in pursuit.<\/p>\n

Why did that happen? ForkLog traced the development of both chatbots: their new features, day-to-day use and the key differences between the assistants today. Beyond numbers and benchmarks, we looked at real-world experience with each system.<\/p>\n

A rocky start: Bard\u2019s first steps<\/strong><\/h2>\n

On the face of it, Bard did not stand out from rivals: the same prompt box and a dialogue-style output window. The notable distinction was its ability to tap Google Search for fresh information on the web.<\/p>\n

\"From
Start screen of the Gemini chatbot (formerly Bard). Data: Google.<\/figcaption><\/figure>\n

From the outset Bard ran into a serious problem: critical inaccuracies in its answers. Other chatbots suffered similar lapses, but at Google the scale was unprecedented. The starkest example\u2014a mistake in Bard\u2019s own promotional materials\u2014knocked $100bn off Alphabet<\/span>\u2019s market value within a day of the announcement.<\/p>\n

The rush was to blame. Google hurried to ship a response to ChatGPT and Microsoft Copilot (which runs on OpenAI\u2019s technology). The firm enlisted its own staff to test the bot, asking them to manually fix incorrect answers to smooth user interactions.<\/p>\n

Matters were worsened by relying on the older LaMDA language model even though the company already had the more advanced PaLM.<\/p>\n

At the time the service was in beta with a limited number of users. The public release arrived in July 2023 and, despite the rough start, Bard built an audience.<\/p>\n

According to data<\/a> from Coolest Gadgets, in 2023 the chatbot\u2019s website drew more than 142m monthly visits. For comparison, ChatGPT\u2019s peak in May that year reached<\/a> 650m.<\/p>\n

The road to refinement: from LaMDA to Gemini 2.0<\/strong><\/h2>\n

Over two years Google\u2019s chatbot travelled far. In April 2023 the company\u2019s CEO Sundar Pichai announced<\/a> Bard\u2019s move to a more powerful model\u2014from LaMDA (137bn parameters) to PaLM 2 (540bn). He said<\/a> the initial choice of LaMDA reflected its modest compute demands, which allowed a quicker launch and faster collection of user feedback.<\/p>\n

The switch to PaLM 2 markedly improved Bard\u2019s capabilities. That is borne out by MMLU<\/a> (Massive Multitask Language Understanding) tests\u2014a standard set of 16,000 questions across 57 disciplines, from maths and physics to law. The benchmark gauges a model\u2019s ability to tackle varied tasks: the higher the share of correct answers, the more versatile the system.<\/p>\n

PaLM 2 scored 78.3%, handily beating GPT-3.5 (70%), which powered ChatGPT in early 2023. But in March OpenAI released GPT-4 with 86.4%, again putting Google behind.<\/p>\n

The Gemini era: a new phase<\/strong><\/h2>\n

The turning point came in December 2023 with the shift to Gemini, built by DeepMind. Three versions\u2014Ultra, Pro and Nano\u2014improved text, code and maths generation. It was Google\u2019s first public multimodal model for analysing images, video and audio\u2014a capability GPT-4 had added back in March 2023.<\/p>\n

In MMLU tests Gemini Ultra narrowed the gap, posting 83.7%. Even so, by May 2024 OpenAI had introduced GPT-4o with 88.7%.<\/p>\n

In February 2024 Gemini 1.5 arrived<\/a> with a Mix of Experts approach, in which specialised networks tackle complex tasks in parts. The context window expanded to 1m tokens\u2014the equivalent of one hour of video, 11 hours of audio or 700,000 words. In December of the same year Gemini 2.0 Flash Experimental was released<\/a> with improved performance and new features, including real-time audio and video interaction.<\/p>\n

Google lavished attention on mobile. In February 2024 it launched an Android app and integration with Google on iOS. Android users could replace Assistant with Gemini, although at first it lacked basics\u2014it could not place calls, control smart-home devices or even set an alarm.<\/p>\n

That changed after the Google I\/O conference in May 2024. Gemini was integrated with Chrome, Photos, Workspace and other services via extensions. The smartphone version now can:<\/p>\n