There’s a new space race heating up, and this time, Microsoft and Google race to plant AI on their search results. Both companies have announced AI technology to power their search results. So what will actually change and which one will be better? Let’s take a look at the features announced so far.
Table of Contents
Artificial Intelligence Crash Course
Since OpenAI’s release of ChatGPT, an AI chatbot, AI has entered the chat. Within a matter of weeks (if not days), AI has entered the world’s knowledge.
Tools powered by AI have been on the market for a while from text generation to image creation, AI is a multi-billion dollar industry.
A quick crash course in AI buzzword:
We’ll go more in-depth on these topics and topics to see how they will change how we search on the internet!
AI-Powered Search Engine Result Pages (SERPs)
For search engines like Google and Microsoft, their goal is to deliver the best answer for a user’s search query.
When you search “best places to travel” both search engines provide results from across the web from travel booking sites, enthusiast blogs, and travel agents.
The new AI systems promise search results based on machine learning to provide answers on complex topics and queries. Search results could become more detailed, accurate, and personable since they are trained in natural language models.
With this new technology, results for “best places to travel” provide opinionated and explained results straight from the search engine rather than a third-party resource.
However, a lot of improvements still need to be made. Since Google and Microsoft announced their AI search results, artificial intelligence has been the talk of the town.
While both companies will start applying AI to their results, the language models behind the AI tools are different.
Microsoft Bing and OpenAI
Microsoft announced they invested with OpenAI in 2019, and is now reaping the rewards. The new Microsoft Bing will use OpenAI’s language model trained GPT-3 with Microsoft’s proprietary AI model Prometheus.
Backed with this new technology, Bing is in a beta test with a limited user base for the AI chatbot. In the coming weeks, Microsoft will introduce the new Bing to all users in hopes of drawing traffic away from Google.
Google is not going out without a fight though. After Microsoft announced the new Bing, Google showcased Bard. Google Bard is a proprietary AI making use of Google’s vast data to produce conversational, high-quality responses.
Google Bard has yet to roll out to select users but Google has debuted example queries that promise new features to the results page.
Language Models Go Awry
So far, for Google Bard and Microsoft Bing it hasn’t been all sunshine and rainbows. During Bing’s beta launch they began limiting user’s search sessions due to repetitive and useless responses from the AI.
Google came under fire when Bard produced the wrong answer to a question about the solar system. When asked about new discoveries from the James Webb Space Telescope, Bard responded with multiple answers including the assertion that the JWST “took the first pictures of a planet outside of our own solar system.”
Unfortunately for Google Bard, the first pictures were taken two decades ago in 2004 by the European Southern Observatory’s Very Large Telescope. This blunder cost Google billions as Alphabet shares dropped 7.7%.
Despite some hiccups, the future of AI search results seems inevitable. More and more AI dialogue applications pop up every day. So, which will change how the world searches?
What Will AI-Powered Search Results Look Like?
Google Bard and Microsoft Bing promise detailed, multi-faceted answers from their AI search engines. Historically, large language models are limited by their data, which can lead to outdated responses and gaps in knowledge.
Both companies promised their search results have proprietary source code that enables their tool to be the most up-to-date response on the web.
In typical Google fashion, the new result seamlessly joins the user interface. The AI result appears similar to other search features like the Knowledge Panel, Featured Snippets, or FAQ results.
That’s good and bad news. While it does replace some valuable space in the SERP. It also has a “read more” feature which highlights results related to the query which are the typical articles, blogs and forums that would’ve appeared.
Google’s language model: LamDA
One of the most exciting elements for AI search is the ability to quickly interpret a wide range of perspectives and synthesize the information for a thoughtful answer. This works the best when it has the most up to date information to work with.
Google used their company web crawlers to develop a data set C4 (Colossal Clean Crawled Corpus). C4 combined with their Natural Language Processing (NLP) model LamDA, provides high quality responses with recent information.
Bing Search Engine
Bing has already launched the new search services to select users and it seems to have already adjusted how it appears. Instead of a taking space from the typical search results, Bing AI appears on the left-hand side.
Once the query loads, Bing AI begins a writing animation which you can stop at any point. For longer responses (and Bing’s responses can be long) the writing takes a while, to the point where it’d be faster to click on one of the blogs that appeared.
Bing’s sample queries are strategic, the queries are detailed enough to highlight the intelligence of the AI chatbot but subjective enough to mask any potential errors. Most common applications will likely be subjective in nature which fits Bing’s language model.
Microsoft GPT-3 and Prometheus
When Microsoft partnered with OpenAI ChatGPT’s parent company, consumers were excited at the prospects. ChatGPT created an AI conversation around the world. However, ChatGPT has other applications, like their new subscription service.
Microsoft needs to create a differentiation between ChatGPT and Bing. After debuting the AI upgrade, Microsoft announced Prometheus, their data model. Prometheus provides context to the search results. Bing AI results will have the details of real-time events like weather, sports, news and more.
Applications of AI in Marketing
The experimental conversational AI service will undoubtedly change the landscape of SERPs. So, to market your product, or service it’s important to make use of any new tools on the market.
ChatGPT has the ability to generate content for marketers from Twitter posts to long form blogs. But, with these AI search engines, we need to expect different use cases.
Google’s C4 uses crawled information from the internet, so it may be time to revisit your website’s crawl budget.
Bing has reentered the chat with their AI chatbot, so turning some marketing power to Microsoft could pay off in the long run.
I tested Bing AI out and it had some impressive results, being able to not only write about marketing but do it in iambic pentameter.
While impressive, so far AI has not replaced typical search results, and until consumers start interacting with the new service, we can’t know for sure how it will change content marketing strategies.
Only time will tell, but you could try asking AI too.
Who Will Get to AI Search Engines First?
It’s a battle to topple the king. Google has dominated as the world’s search engine, but this new Bing might be enough to dethrone them. The race for AI dominance is just heating up. But one thing is for sure: both Google and Microsoft are innovating the way people explore content.
A business needs to closely watch emerging technologies and changes since they have the potential to make a huge difference in SEO and keyword strategies. Rescue Marketing keeps a close eye on exciting developments to make use of cutting-edge techniques and tools.
Follow our marketing experts — we’ve got you covered!