This week in AI, Google announces a major expansion for its Bard AI platform and hypes the coming release of its multimodal Gemini software.
It seems like it was just yesterday when Google provided us with its first look at its Bard AI platform with the infamous factual errors. That blemished presentation has tainted many people's view of Google as a true competitor in the AI space. Shares of Alphabet fell by nearly 10% following the event, and many attributed this to another failed opportunity by Google's management team.
Not to be a ~well, actually~ jerk, and I'm sure Bard will be impressive, but for the record: JWST did not take "the very first image of a planet outside our solar system".
— Grant Tremblay (@astrogrant) February 7, 2023
the first image was instead done by Chauvin et al. (2004) with the VLT/NACO using adaptive optics. https://t.co/bSBb5TOeUW pic.twitter.com/KnrZ1SSz7h
Despite this early hiccup, earlier this week, Google announced an expansive upgrade to that same Bard AI platform with new integration with the full suite of Google products and services. These include YouTube, Google Docs, Google Drive, and its popular travel site, Google Flights.
We’re adding extensions to Bard so you can connect it to your favorite Google apps including Gmail, Drive + Docs for even deeper collaboration. We’re also updating how we validate the claims in Bard’s responses with an improved “Google It” button + more.https://t.co/whwuXoicS2 pic.twitter.com/63oEtucxTZ
— Sundar Pichai (@sundarpichai) September 19, 2023
Bard AI is Google's answer to OpenAI's large language model, ChatGPT. Other enhancements with this update include support for multiple languages and more robust fact-checking. It is the largest update for Bard AI since it was released in March.
The enhancement to Google Flights looks to be one of the more significant upgrades. Not only will Bard AI be able to provide integration between Google Maps and hotel searches to provide users with the most efficient routes and itineraries for their travel planning.
Google's full suite of products and services is one of its main competitive advantages over other companies. Its Google search engine remains the global leader in internet browsing, while other products like Gmail and Google Docs are used by billions of people each day.
One unique feature that Google is test-driving with Bard AI's latest update is the "Double Check" button. This button will allow users to check the accuracy of Bard's responses and outputs. The idea here is that if something sounds questionable to a user, they can essentially run a back-test to see if Bard is providing a hallucination. These hallucinations are when an AI chatbot will confidently provide a factually incorrect response.
Another feature that Bard is introducing is the ability to share your personal conversation with a friend. That friend can then continue on with Bard AI with the same stream of consciousness from the original conversation.
Bard AI has come a long way since it was rushed for release to try and keep up with OpenAI. Since then, Google has taken time to refine the platform and now that it is rolling out full-scale integration with its services, it could be the AI name to back moving forward.
Google Teases Imminent Release of Google Gemini AI Software
It wasn't just the Bard AI upgrade that Google announced this week. The tech giant also selected a handful of companies to be early users of its Gemini AI software. Gemini is the engine behind Google's various chatbots which includes Bard AI. It is a direct competitor to OpenAI's GPT-4 platform, although the version that is being tested right now does not have Gemini's full capabilities.
Why are people excited about the release of Gemini? It is the brainchild of the AI experts at DeepMind and Google's Brain Team. Gemini will be a multimodal system which can integrate text, images, and other types of data. At the Google I/O developer conference in May, CEO Sundar Pichai also intimated that Gemini will be able to handle tasks that require deep reasoning in the future.
Gemini is built on DeepMind's AlphaGo platform which has drawn accolades based on its ability to master the game of Go. It will also be integrated with Google Pathways to enable scaling and a diverse range of training from large sizes of datasets. This means that Gemini has the potential to be a much larger language model than GPT-3.
Google isn't the only big tech company that is racing to compete with OpenAI. In fact, major OpenAI stakeholder Microsoft is teaming up with Meta Platforms to power its recent release of its Llama 2 LLM.
The current testing phase of Gemini promises that Google is close to a full-scale launch of the platform. While there are no guarantees in the tech world, Google's Gemini has the opportunity to disrupt the generative AI industry as we know it. Pichai has stated that Gemini will likely operate as a virtual assistant in the future, especially as it becomes more integrated with Google's services. It will eventually be able to help the user in both their personal and professional lives.
There is no hard launch date for Gemini with some sources citing both October and December as potential releases. Either way, all eyes will be on Gemini when it is released as the world gets its hands on what will likely be the biggest challenger to OpenAI's GPT.
