ChatGPT Won't Kill Google. It Will Help It.
And the bigger AI race you should pay attention to
Microsoft 🤝 OpenAI
Last week, The Information reported that Microsoft and OpenAI are working on a ChatGPT-powered version of Bing.
Since ChatGPT was released, it has been used for countless tasks, from debugging code and writing essays, to creating languages and inventing Seinfeld scenes.
Another use case that raised eyebrows was ChatGPT's ability to answer questions in concise and straightforward language. It seemed like a much better experience than the one we’re used to on Google:
Type your search query
Scroll past ads (which are getting harder to differentiate from non-ads)
Skim through the search results/links
Make your best guess on which one will answer your question
If you didn’t get your answer, go to Step 3 and repeat
This only bothered us once ChatGPT came along and showed us a different experience. ChatGPT spoonfeeds us the information, while Google gives us a menu to choose from.
Of course, Google does occasionally provide similar answers without having to click on any links:
Still, ChatGPT was alarming enough for Google to declare a "code red" and prioritize the release of its own AI products. Many were quick to declare the death of Google, and I couldn’t disagree more.
How Google will benefit from this
ChatGPT is a threat that Google shouldn't take lightly. When Microsoft and OpenAI release a ChatGPT-powered search engine, many of us will flock to it (myself included).
ChatGPT is currently trained on data until September 2021 and isn't connected to the internet. But it will be once it integrates with Bing's search engine. This opens up a realm of risk and uncertainty, including providing false information.
Microsoft and OpenAI will do whatever they can to anticipate and minimize those. Still, nothing is guaranteed for a first-of-its-kind product like this one.
This doesn't mean that a ChatGPT-powered Bing will fail, just that Microsoft is taking a massive reputational risk. And it's understandable. Currently, Google commands 85% of the global search market, while Bing only accounts for 9%. So Microsoft doesn’t have as much to lose as Google, which explains its higher risk tolerance. And it wants to reap the benefits of the $1 billion it invested in OpenAI.
Google has been building similar large language models (LLMs) but has been much more secretive and cautious about them. You may recall last summer's controversy when a Google engineer claimed that LaMDA, Google's chatbot, was sentient. And Google even built one of the core technologies powering ChatGPT. We don’t know how powerful Google’s chatbot and other AI products they’ve created are, but we know they have the talent and funds to compete.
Search ads account for about 60% of Google's revenue, and they'll be careful releasing any products that threaten their biggest money-maker.
By being the first mover, Microsoft’s chatbot search engine will help Google. It will shed light on the questions and concerns that have kept LaMDA under wrap. Google will closely observe how people use (and misuse) Microsoft’s chatbot, how much they're willing to pay for it, and attempt to create a better version.
The first mover takes the most risk and doesn't always win. Even Google Search wasn't the first search engine, but it became the most successful.
If Microsoft’s new search engine succeeds, Google still faces the innovator's dilemma: how can they offer a chatbot search engine without cannibalizing Google Search?
The most interesting AI race isn’t in Search
Search isn’t the most interesting part of the story. We're approaching an inflection point that will reshape the information age.
In another report earlier this week, Microsoft was looking into integrating OpenAI's chatbot tech into Word, Excel, PowerPoint, Outlook, and other Office apps.
What excites me the most about ChatGPT is its ability to generate content based on a prompt rather than using it as a search tool, which is why I find this more significant than the Bing integration.
Imagine you own a company that uses Microsoft Office products. You could pay extra to get ChatGPT-like functionalities into Word, Excel, PowerPoint, and Outlook. This would boost employee productivity exponentially: reports, spreadsheets, and presentations could be made in a fraction of the time and require a few final touches.
This becomes a significant competitive advantage for your company against your competition. Meanwhile, FOMO will kick in for your competitors, and they'll also decide to get the premium tier of Microsoft Office so their employees can be just as productive.
23% of Microsoft's revenue comes from Office products, compared to the 6% from ads on Bing. So even this is bigger news for Microsoft as it cements its #1 spot for office/productivity applications.
Google's GSuite also competes with Microsoft Office. And now GSuite will seem outdated if Google doesn't offer the same generative AI functionalities in Gmail, Docs, Sheets, and Slides. Another reason that justifies Google's code red.
So the competition extends to Microsoft and Google's biggest products:
Search: Bing vs Google Search
Documents: Word vs Google Docs
Presentations: PowerPoint vs Google Slides
Spreadsheets: Excel vs Google Sheets
Email: Outlook vs Gmail
The office apps that have remained relatively stagnant for the past two decades are about to be flipped over their head.
We're at a significant inflection point. What a time to be alive.
Poll: Bing vs Google chatbot
Let’s fast forward to the near future when Microsoft Bing and Google are both chatbot-powered search engines. Which one would you use?
As always, the comments are open for discussion.
Deep dive
If you enjoyed today’s story, I compiled my favourite stories about ChatGPT and the future of AI that I’ve encountered in the past few weeks:
What Happened To Amazon’s Employees After AI Automated Their Work on
Not ChatGPT-related but another great perspective on the future of AI. The common narrative (and fear) around AI is massive job loss/replacement. In this article, Alex goes behind the scenes at Amazon and shows us how automation is a natural process of evolution and will redefine jobs rather than replace them. This is a perspective I agree with: AI will automate specific tasks rather than our entire jobs.
Generative AI: autocomplete for everything on
An optimistic perspective on how generative AI will impact the future of work. Noah and roon share specific examples of how writers, programmers, designers, and others will benefit from AI as a “productivity-enhancing” tool. My favourite part of the story is the “comparative advantage” argument.
Generative AI: My Enhancement, Your Replacement on
This is 's counter-argument to the article above. He argues that AI could lead to partial replacement since AI’s enhancing abilities will require fewer humans in the long term. I loved reading this because it exposed the gaps in my arguments and perspective on the future of AI.
The Hidden Cost of Artificial Intelligence (Jack Raines)
Since ChatGPT was released, I’ve been torn about AI’s impact on us. I liked Jack Raines’ piece on this. He says: “When you outsource writing to AI for the sake of speed, you sacrifice the development of your own thoughts and the exploration of your own opinions”. I hope generative AI tools will be treated more as an accelerator for ideation and brainstorming rather than becoming tools we fully depend on from start to finish.
AI for the Next Era, an interview with OpenAI CEO Sam Altman
OpenAI CEO Sam Altman sits down with Reid Hoffman (co-founder of Linkedin and partner at VC firm Greylock) to talk about the future of AI.
Thanks for reading Year 2049! This article is free for everyone to read, and you can support me by sharing it with others.
Tool: There’s An AI For That
While we’re on the topic of AI tools, I found this useful repository that categorizes them by use case. It’s called There’s An AI For That.
2023 Goals
The winter break gave me much-needed time to reflect on what I want to achieve with Year 2049 this year. It’s less about numbers and more about content:
Experiments: Instead of just talking about new technologies, I wanted to put my technical and design skills to good use and start sharing experiments I’ve been doing.
Interviews: I’ve written about many fascinating companies, but I'll start reaching out to them and bring you exclusive interviews rather than me quoting things I’ve read.
Formats: The newsletter’s format has been pretty much the same in the 44 editions I’ve sent so far, and I’ll explore new ones that will be shorter but just as thought-provoking.
The last goal is to have more fun, which is something I lost sight of in the later stages of 2022. Because if it’s not fun, what’s the point?
See you next week,
Fawzi
How would you rate this week's edition?
The whole AI field is concerning to me. I don’t like software having the ability to “think.” It is already hurting creative types like illustrators. I imagine many more jobs will be lost in several fields. It will probably write web copy and ads eventually. And of course, we know that what was once science fiction can become science fact. Satellites for example. IPhones, etc. Those old movies about robot takeovers may end up being true. Machine have no moral compass (of course, neither do most large corporations.) 😉
Also, really? Justine Moore bringing up Clippy after all these years? May it rest in peace and never return—IMHO—Clippy was an obnoxious pain and the world is safer now.