Google’s Bard vs ChatGPT: Inside the New AI Chatbot and Its LaMDA Roots

Google’s release is aimed directly at ChatGPT, and although full completion will take a few weeks, the product is already in a testing phase.

On February 7, Sundar Pichai surprised many by announcing Google’s next‑generation conversational AI system called Bard.

Since the explosive rise of ChatGPT, major tech companies have accelerated AI development, and Google views ChatGPT as a serious competitor, assigning it a “code red” priority.

ChatGPT, launched on November 30 last year, gained over one million users in five days, prompting established tech giants to reconsider their AI strategies.

Within Google, an internal chatbot named “Apprentice Bard” has been undergoing response testing similar to ChatGPT.

Apprentice Bard is built on Google’s LaMDA technology, analogous to how ChatGPT is based on OpenAI’s GPT series. LaMDA, developed by Google Research in 2017, stands for “Language Model for Dialogue Applications.”

Google has tasked the LaMDA team with focusing on countering ChatGPT’s expansion. Bard’s public test version derives from Apprentice Bard.

Bard’s interface is user‑friendly

Bard operates in a chat‑box format like ChatGPT, capable of simplifying complex topics (e.g., explaining the James Webb Telescope to a nine‑year‑old) and assisting with tasks such as planning a baby shower, comparing Oscar‑nominated movies, or suggesting lunch ideas based on fridge contents.

Google describes Bard as an experimental conversational AI service powered by LaMDA, aiming to combine the breadth of world knowledge with the power, intelligence, and creativity of large language models, delivering up‑to‑date, high‑quality responses sourced from the web.

LaMDA is a statistical model that predicts the next word in a sequence, enabling more natural and fluid dialogues compared to task‑specific AI models. Earlier LaMDA versions featured up to 137 billion parameters, approaching human‑level conversational quality.

The initial Bard release will use a lightweight LaMDA model requiring less compute, allowing broader user access while the team—largely composed of Chinese engineers—continues to refine functionality.

Google’s public test aims to gather feedback to improve Bard’s quality, speed, safety, and grounding in real‑world information.

Key Differences Between Apprentice Bard and ChatGPT

Provides real‑time responses using up‑to‑date web information, unlike ChatGPT which is limited to data before 2021.

When using Google Search, users can switch to Apprentice Bard and retain the same search results; ChatGPT replies are text‑only.

Google’s Bard is built on LaMDA, while ChatGPT is built on OpenAI’s GPT technology.

ChatGPT includes an AI text classifier for generation detection; Bard currently does not.

Apprentice Bard is free to use, whereas ChatGPT offers paid subscription tiers.

Google emphasizes a bold and responsible approach to AI, having published AI principles in 2018 and continuing to collaborate with governments and external experts to ensure safe and useful AI.

Beyond Bard, Google is integrating AI advances such as BERT, MUM, PaLM, Imagen, and MusicLM into its products, starting with search and expanding to other services.

Developers will soon be able to access language‑generation APIs initially powered by LaMDA, with further models to follow, supported by Google Cloud partnerships.

In summary, Bard represents Google’s answer to ChatGPT, leveraging LaMDA for conversational capabilities, real‑time web integration, and a free access model during its early testing phase.