Will AI Powered Microsoft Bing Kill Google in Search?

Microsoft Bing has about 3.2% of the overall global search engine market but has 8-9% of the desktop search market.

It took Google about two years after getting to scale to pass Yahoo in 2002 as the top search engine. Microsoft Bing already has some global scale. It will take about 500% growth on the desktop for Microsoft Bing to get the majority of desktop search and it would take about 1600% growth to pass Google in mobile search.

Microsoft stated that the AI enhancements increased search relevance by the largest improvement in 20 years. Twenty years is when Google won search with improved search relevance. Google will fight back and create its own AI-enhanced search engine. If there is search relevance parity or a period of leap-frogging search relevance then we can still have a condition where search engine market share becomes competitive. Microsoft and Google could battle at 30-50% market share for a few years.

We will see AI chat-powered search engine wars for the next two years.

Multiple potential players have the resources to compete in new AI-powered search, social media and video. All of the old internet business moats are not safe. Facebook is not safe with social media. Twitter will need to make its own AI-powered services.

Meta’s version of the GPT is called the Open Pretrained Transformer, or OPT. Meta has two other Large Language Models (LLM) one called Sphere and the Galactica.

Google has a new chat system called BARD.

Baidu has Erniebot.

There are also many new AI startups.

More people use Google on their mobile devices than on desktop computers (95% versus 85%).

In November, 2022, Microsoft announced plans to double its search engine (and news) advertising from about $10 billion per year to $20 billion.

Microsoft has radically improved Bing by adding an OpenAI AI co-pilot based off of ChatGPT technology. There is still a waitlist to get access to the chat.

8 thoughts on “Will AI Powered Microsoft Bing Kill Google in Search?”

  1. I think there is so much clutter and marketing on web it is often hard to find good data. I used to find good data using google way faster in the past than now.

    Perhaps the reason is that quality is not so important than marketed high ratings on search engine or search engines, Facebook is designed in a way that keeps people at the screen as long as possible. That is because if you spend more time on the web, you see more advertising and they earn more money. So they want to keep you on web longer.

    As for Chat GPT it seems really fast and scarry good at some things. The results of chat cpt search can be false, and the essay or summary just not true. Free thinking is more important than ever.

  2. Microsoft Bing is already better than Google, if you’re into serious research and not just using it to find restaurants or sneakers (Yelp or Amazon might be better for these specific purposes, respectively, anyway).
    The Correct-Think objections raised here are serious concerns, though I think Truthfulness is a bigger one. Like a lot of human Blowhards, ChatGPT can be awfully confident, even when it is wrong.
    I asked another AI engine, Lexi.ai about my 88 million pound prospective building; its own estimate – because ChatGPT didn’t know about it, and Lexi.ai did. I asked it how deep the foundations should be. It confidently spit back ~12″. My engineer collaborator and I had a good belly laugh about that one. Meanwhile two structural engineers familiar with the bedrock of the proposed sites said the foundations would have to be 175′-200′ deep. No one would want to build anything real from an AI powered engine. And policy papers or health regimes etc. are probably just as reliable.
    But, it’s a great start, and fine for basic research, as long as you check the sources. Maybe the best thing about these tools is the ability to help semi-literate people write coherent texts (though this screed here just too me ~10 minutes).

    • To be fair, I’d fire any engineer who came back with a foundation depth that they got from ANY internet search engine. The fact that it’s AI is neither better nor worse.

  3. ChatGPT is beyond scary. I asked it to write me a 1500 word essay on a random topic (importance of vegetables) and it blew me away with the write up it produced. I mean this is some seriously unbelievable stuff.

    • Are you aware that ChatGPT can just as dazzlingly produce a result that is factually wrong as one that is factually correct? This has been demonstrated repeatedly. It has no way to know whether the words it has found to string together make correct statements or incorrect statements. It only knows that they are grammatically correct.

      That should be more scary to you than being impressed at the essay it wrote about vegetables. (pardon me if I misunderstood the thrust of your comment)

  4. Microsoft wants you to give them access to your psychological switchboard.

    How do you think they want to re-route your thoughts and beliefs?

    To make you a better thinker, or, to make you think “better”?

    They talk about reducing cultural biases –based on their definitions of what a “better” culture is?

    They talked about, not just ignoring problematic searches and questions, but, identifying them, and reacting “appropriately”.

    Turning people in to the government for violating “better-think” standards, as Microsoft sees fit to define them?

    A true personal Agent, owned and controlled by individuals, is a far better path.

    If I owned the trained model, as a separate and isolated Agent, that accessed the raw data, then I might trust that the answers I got would be closer to what I wanted, without having been tuned to “improve” me.

    Unless I’m asking for my Agent for counseling.

    • That is precisely the problem here. Maybe 20-30 years ago, this would have been fine. But today?

      The culture in high tech is incredibly controlling, and dominated by a particular political perspective, and it is utterly and completely predictable that an AI assistant running on some platform’s hardware is NOT going to be working for you.

      It’s going to be manipulating and watching you, and compiling a social credit score on you, and it WILL eventually report you to a hostile government agency if it suspects you of wrong think, if not just deal with you directly.

      We can’t evaluate these things as though we didn’t know what’s been going on with all the major platforms. They’re an embryonic police state every dictator in history would have envied.

    • It will take a long time for Microsoft to override Google as the premier search engine, not because it isn’t better, but because the advertising is wrapped around Google like a coiled snake.

      I agree with you that we shouldn’t trust Microsoft to control and direct what they deem the correct thoughts to be typing into your keyboard.

      I can’t even trust them to make a video game controller that doesn’t stop working in a month so you have to shell out another $70.00.

Comments are closed.