Background for sudden introduction of Bard without much preparation.
Google has been the undisputed king of web searches for more than two decades, with a net worth of 1.15 trillion dollars in 2022. However, ChatGPT, a conversational chatbot, has become a technology sensation and a potential threat to Google’s central business model.
The alliance between Microsoft and OpenAI started when Microsoft invested one billion dollars into OpenAI, and Microsoft was able to acquire OpenAI’s intellectual property portfolio in the process. Since then, Microsoft and OpenAI have been working together on various projects, and specifically on integrating ChatGPT into Bing.
New artificial intelligence (AI) powered Bing and Edge was introduced as “The new AI powered Bing and Edge – Your copilot for the web”. Satya Nadella, Chairman and CEO, Microsoft showed up highlighting a lot of new features mainly,
A new chat experience – in the demo he did a complex search for a gaming monitor with preferred budget by switching to the chat mode, which was a totally different experience in a search engine.
A creative spark – when you need more than just a search, such as when you need to get some inspirational content, the new bing can help you with it by generating the content.
New Edge with new look and features – Alternative to google chrome, along with Bing, Satya Nadella introduced new Edge with new look and functionalities like,
With the Edge Sidebar, you can ask for a summary of a lengthy financial report to get the key takeaways. You don’t need to read any report.
More than that it can search on the internet some other report and compare them in a table for you.
Is new Bing and Edge really a threat to Google business model?
Google dominates 90% of search queries, while Microsoft’s Bing only accounts for 9% of worldwide desktop searches in 2022. Chat GPT is a potential existential threat to Google’s business because Google relies on search for 60% of its revenue, and on the other hand Chat GPT offers an alternative to search that is powered by AI.
These moves from Microsoft side has made Google panic, and an emergency code red situation was declared and their competitive alternative to ChatGPT, Bard was introduced without much preparation. This made the situation worse.
Google shared an example of Bard answering the question “What new discoveries from the James Webb Space Telescope can I tell my 9 year old about?”. The ad included an incorrect claim that the James Webb Space Telescope took the first pictures of exoplanets. Social media users pointed out the error, and the company’s shares slid by as much as 9% during trading hours. The ad aired just before senior executives touted Bard as the future of the company at a launch event in Paris.
Is OpenAI chatGPT much powerful than Bard?
In 2021, Google announced their own breakthrough conversation technology, called LaMDA (“Language Model for Dialogue Applications”) at Google’s annual I/O event. Bard is Google’s AI chatbot powered by LaMDA.
ChatGPT vs Bard structural differences.
Both are language models specifically, “Generative AI models” which create content after a user queries it. Chat GPT to the date is a trained model which generates answers for the queries with its trained data which was available before 2021. On the other hand, Bard provides upto date responses fetching from the web without depending on training data set.
In general, generative language models challenges:
High Cost per generation/ prediction:For example: The analysis reveals that ChatGPT is deployed on Microsoft’s Azure cloud. Thus, there is no need for OpenAI to invest in a physical server room but however currently,
Microsoft charges $3 per hour for a single A100 GPU, and each word generated by ChatGPT has an associated cost of $0.0003.
Assuming a response from ChatGPT will usually have at least 30 words, the cost of a single response generated by ChatGPT is at least 1 cent. It is estimated that OpenAI has a running cost of at least $36 million per year.
Privacy preserving with the collected sensitive Data entries is challenging.
Lack of high quality data availability and much labor intensive refining. For an example to avoid answering queries which threaten the well-being of the human and society.