Google's search chief warns against solely relying on AI chatbots for accurate information

The head of Google's search division, Prabhakar Raghavan, has warned against relying solely on AI chatbots to provide accurate information. He stated that these chatbots can sometimes generate false answers that appear to be truthful. Raghavan referred to this as "hallucination," where a machine produces a plausible but entirely fake response, as per a report from Sky news. Despite this, Google recently unveiled its own AI chatbot named Bard, which they hope will compete with ChatGPT. However, an advertisement for Bard showed it providing an incorrect answer to a question about the James Webb Space Telescope.

I think the Google presentation signaled (rightfully or wrongly) that they are not really taking it seriously: it was less than 5 minutes, it was presented amongst other feature, it was in Paris instead of HQ, CEO wasn't there. Bing, on the other hand, changed the homepage.

— Maarten Bosma (@MaartenBosma)

Shares of Alphabet, the parent company of Google, plummeted due to the recent error and growing concerns about the potential threat posed by ChatGPT to Google's search dominance. Raghavan has emphasized the importance of releasing Bard to the public but also acknowledged the great responsibility that comes with it. The company is exploring how to integrate AI capabilities into its search functions, especially for questions that have multiple answers. The comments come after some criticized Google and CEO Sundar Pichai for the poorly executed launch of Bard. Maarten Bosma, a former research engineer at Google Brain, pointed out that the presentation demonstrated the company's lack of commitment to AI.

What is AI and how does it work?

Artificial intelligence (AI) is a branch of computer science that aims to create intelligent machines that can perform tasks that typically require human intelligence, such as visual perception, speech recognition, decision-making, and language translation.

There are several different approaches to AI, but the most common methods include:

Machine learning: This involves feeding large amounts of data into algorithms, allowing the algorithms to learn patterns and make predictions or decisions based on that data. There are different types of machine learning, including supervised learning, unsupervised learning, and reinforcement learning.

Natural language processing (NLP): This involves enabling machines to understand, interpret, and generate human language. NLP is used in applications like language translation, text-to-speech, and sentiment analysis.

Robotics: This involves the development of robots that can perform physical tasks and interact with the world in a human-like way. Robotics often involves combining AI techniques with engineering and physical design.

Deep learning: This is a subfield of machine learning that is inspired by the structure and function of the brain, called artificial neural networks. Deep learning algorithms can be used for tasks like image and speech recognition.

At a high level, AI works by allowing machines to learn from data, make predictions or decisions based on that learning, and continually improve as they are exposed to more data. The success of AI depends on the quality of the data, the algorithms used, and the computational resources available to run those algorithms.