Google opened ChatGPT competitor Bard for the public Tuesday in the United States as well as the United Kingdom. However, you will need to wait on a waiting list.
“Today, we’re opening access to Bard,” Google Vice President of Product Sissie Hao and Vice President of Research Eli Collins stated in a blog post.
They said that Bard could be used in order to increase productivity, speed up the generation of new ideas, and spark curiosity.
They noted that they have learned a lot by testing Bard, and the next step is to get feedback from more people.
The executives at Google acknowledged that large language models can be a fascinating technology. However, they have their flaws. They learn from a variety of information that reflects real world biases and stereotypes. This can sometimes lead to errors in their outputs, the executives said. They can also present false, misleading or inaccurate information confidently, while still being able to provide accurate, relevant, and convincing information.
The pair said that “our work on Bard is guided primarily by our AI Principles and that we continue to place emphasis on safety and quality.” We’re using human feedback to improve our systems and have also implemented guardrails like limiting the number of dialogues per day to ensure that interactions are helpful and on-topic.
Google has been trying to catch up to Microsoft since February when it unveiled Bard to all the world. Microsoft has been adding AI features to its products at an alarming pace ever since.
“Google is panicking now that Microsoft beat them to the market and they are bleeding users which puts them into a’ready-or-not here it comes’ mindset,” said Rob Enderle who is president and principal analyst at the Enderle Group advisory services firm in Bend.
He said that Microsoft had convinced them it wasn’t ready a while back and reduced resources to it. It is therefore unlikely it is ready. However, they need to respond now and clearly take a huge risk to stop Microsoft from draining their search engine.” TechNewsWorld.
Google is undoubtedly in catch-up mode, stated Mark N. Vena (president and principal analyst at SmartTech Research, San Jose, Calif.).
He told TechNewsWorld that he believes Google is under immense market pressure to get Bard into mainstream market ASAP because there has been a perception they were caught off guard by ChatGPT’s market reception.
Google has been talking for years about its AI and machine learning work, but it has only reached the consumer in very limited ways up until now, according to Ross Rubin, principal analyst at Reticle Research in New York City.
“ChatGPT really hit a nerve within Google. He said that it was an existential threat for Google Search.
Despite Microsoft’s pressure to close the gap, Bard’s rate of development is still slow.
“Google’s pace seems a little slower than Microsoft’s,” said Greg Sterling, cofounder of Near Media. This is a news, commentary and analysis website.
TechNewsWorld’s he said that Bard feels like it has more to lose as an brand if Bard becomes widely available and goes off-the-rails.
Rubin explained that Bard was being slowly launched because Google holds a dominant market position. He also wanted to position the chatbot in the same way as its existing search products.
He said that Microsoft had a similar rollout using AI within Office.
Vena pointed out that the perception that Google is trailing Microsoft has already caused damage. Google should use its resources to make Bard the most powerful tool on the market, and not worry about being the first.
ChatGPT vs. LaMDA
Vena said that creating a waitlist and slowing down the full rollout can be a benefit to Bard.
He said that it reinforced the perception that Bard was not ready for prime-time. It’s a smart move by Google, and allows them to fix bugs in a deliberate and measured manner.
Sterling stated that the waitlist also serves another purpose. He stated that the waitlist is used to limit access and control who can access Bard. This is how tech products are often rolled out, but it’s fair.
Collins and Hsiao noted that Bard currently uses a lightweight, optimized version of Google’s research large-language model LaMDA. However, the offering will continue to be updated with more powerful models in the future.
“Bard isn’t as powerful as GPT-4, released by OpenAI recently, however, because it’s connected to the internet, this makes a difference in the resources it can depend upon to answer questions,” stated Will Duffield, a policy analyst from the Cato Institute, an Washington, D.C. think tank.
TechNewsWorld’s he said that Bard works better as a personal assistant, but not as well as deep analytical tasks such as asking it for a set of video game patch notes and asking how they will affect the state of play, or asking it to analyze a Supreme Court transcript.
Multiple Answers to Your Queries
Vena explained that LaMDA was specifically created for natural language conversations. It aims to be context-aware and more intelligent than other language models. It was trained on a variety of topics and could be used in different conversational applications such as voice assistants, chatbots, and customer service tools.
He said that Microsoft’s large-language model was not designed specifically for dialog applications, but for general language understanding. Microsoft has been developing several language models to enhance natural language processing across a range of applications including translation, sentiment analysis and question-answering.
Bard also breaks with ChatGPT by creating multiple drafts of its answers to queries. Vena stated that it gives users more flexibility to view multiple queries results and that is a good thing.
Sterling pointed out that multiple drafts offer consumers more information and choice, but it may also be used to protect against user criticism.
He said that Google was being more careful about the messaging and presentation of Bard than Microsoft about Bing.
He said, “Bing is bolder.” “Microsoft is more likely to win than lose, and is ready to embrace AI chat as an evolution to Bing.”
He said, “For Google, this seems to be more a novel addon that will improve with the time.” It’s not being used as a replacement for search. This is partly to influence market perceptions and user expectations.