I’ve harnessed the future of AI web search – that’s why it’s so amazing

The aggressiveness with which artificial intelligence (AI) has moved from the realm of theoretical power to real, consumer-ready products is astounding. For several years, and up until a few months ago when OpenAI’s ChatGPT launched, companies from titans Microsoft and Google to countless startups have embraced the benefits of AI with little practical application of the technology. Everyone knew AI was a thing, but most didn’t really use it.

Just weeks after announcing an investment in OpenAI, Microsoft has launched a public-facing beta of its Bing search engine and Edge browser, built on the same technology that has made ChatGPT the talk of the town. ChatGPT itself was fun to play with, but the introduction of something much more powerful and fully integrated into consumer products like Bing and Edge is a whole new level of exposure for this technology. The importance of this step cannot be overstated.

ChatGPT felt like a toy; Applying the same AI power to a constantly updated search database changes the game.

Microsoft was kind enough to give me full access to the new AI “Copilot” in Bing. It only takes a few hands-on minutes to understand why Microsoft (and seemingly every other tech company) is excited about AI. Asking the new Bing open questions to plan a vacation, create a meal plan for a week, or start researching buying a new TV and letting the AI ​​guide you to something useful is very powerful. Any time you have a question that would normally require pulling information from multiple sources, instantly streamline the process and save time by using the new Bing.

Let AI do the work for you

Not everyone wants to show up on Google or Bing to roll up their sleeves and embark on a multi-hour research session with lots of open tabs, bookmarks, and plenty of reading material. Sometimes you just want to explore a little and have the information delivered to you – AI handles that beautifully. Ask a multi-faceted question and it pulls the information from the web, aggregates it and makes it available to you in a text box. If it’s not quite right, you can ask contextual follow-up questions to generate more fine-tuned results.

Information overload maybe, but this is a great answer.

The biggest thing Microsoft is bringing to the table that wasn’t available through ChatGPT is up-to-date information – and that differentiator is really game changing. While ChatGPT uses a fixed data set and points to a snapshot of the Internet a few years ago, Bing constantly pulls data and always has up-to-date information. After all, it is a search engine that searches the entire Internet. I can get restaurant recommendations for a new eatery that opened last week, information about the new OnePlus 11 that launched yesterday, and current prices for flights and hotels. That alone takes this experience from “fun toy” to “really useful.”

Not only does this make the answers you get from Bing’s AI copilot more relevant, but it also gives the system far more information to work with and build on. If the AI ​​is constantly fed with new data, it can (theoretically) improve faster. And it gets even better as it learns how people interact with it.

But at this point, Conversational AI is far from applicable to every type of search. Or frankly, most searches. And when it doesn’t work, it really falls flat.

This beats ChatGPT… but still has its weaknesses

Flaws we all experience when using ChatGPT aren’t that easy to dismiss or forgive when a company the size of Microsoft puts its name behind it and says it’s ready for everyone. And it only took minutes to test the new AI-powered Bing (admittedly an early beta) to find similar flaws.

It’s counterintuitive, but the Bing copilot works better the less specific your questions are.

Interestingly, the more specific you are in your search, or the more certain you are of the expected response, the worse the copilot’s response. In contrast, it’s more likely to give you a useful answer if you give it less information to work with and leave the question open for the AI ​​to fill in the blanks.

When you ask a very specific question, you often get a very flowery answer with tons of unnecessary context and background. The desired answer is usually available in those few paragraphs and/or several bullet points, and in some cases it’s in bold, but it’s easy to spot when the AI ​​is trying to cover its lack of “knowledge” by writing out a verbose answer . This extra fluff is often not accurate either, as the AI ​​draws from a larger number of sources that tend to be outdated or unrelated.

The week before the Super Bowl when we know for sure who’s playing in the game.

According to Microsoft, over half of web searches are “navigational” or “informational” — meaning I have a specific question about how to get from A to B, or a knowledge base query that’s available in a structured data format. When I ask when a restaurant is open or how many square miles the island of Manhattan is, I don’t need a history lesson coupled with a two-word answer. I just want the answer and move on.

Knowing when to use AI and when to keep it simple

For this half (or more) of searches, the new AI power currently offers little added value. And in many cases it is even an obstacle. Speaking to Bing VP Jordi Ribas, I got some much-needed context on how Microsoft is thinking about this integration of AI into their “traditional” search product and how they manage to bring powerful new technologies to their customers when they knowing that this is not the case is not applicable in every scenario and needs improvement.

Sometimes I just want a simple answer and the AI ​​wants to tell me a story instead.

That debate was strong within the Bing team, Ribas said; There was a strong group of people who felt that moving the entire Bing interface to a “chat” window was the only way, while others were pessimistic, preferring it to be a smaller part of the experience. The result is a very Microsoft approach somewhere in the middle.

When I got my hands on a developer build of Edge with the built-in copilot to use outside of a demo environment, I began to understand what they were getting at. The standard Bing interface is about to start; Except that you now have a larger text box that can hold up to 1,000 characters. If the AI ​​can generate a detailed answer to your query, you’ll get a box on the right with that answer – but it doesn’t inherit the experience, instead supplementing the default search results on the left. If you want to dive in with more questions, you can switch to a full-screen, all-chat experience and have some sort of conversation with the AI. It’s easy to alternately scroll to chat and back to regular results.

And in some cases, at least in this early version, you might not get any AI answers at all – just a standard set of blue links. Microsoft is honest about the fact that these AI answers are very computationally intensive and expensive to return compared to a “normal” search result (servers aren’t free!). So if Bing is confident it can provide a satisfactory answer using the regular search model instead of the new “Prometheus” (AI) model, it will. Although in my experience, as I noted above, it doesn’t fall back nearly enough on this standard model.

Multiple sources, nicely aggregated and displayed with context.

It’s clear that what Microsoft has been able to achieve here is more than just sprinkling a little AI onto Bing. But it’s not that far beyond that, and it doesn’t take long to figure out that at this current level of technology, we don’t want AI in every aspect of our search experience.

The core problem with AI-powered Bing is that the queries, which are demonstrated very well, aren’t typically what you do most of the time when you sit down and use a search engine. I won’t be using AI for a majority of my searches, and at this point the “intelligence” often gets in my way as often as it helps. That takes away some of the sharpness of the astonishment at this announcement.

But I’m still incredibly excited about these developments in AI, and even if you realize you won’t be using AI for every search query, the times you use it can be an extraordinary experience.

This is the future of web search, no doubt about it. It’s just not necessarily the near future.

Editor’s Recommendations