Web search upended
The AI chatbot ‘wars’ have begun. The ways we use the internet will be transformed in ways that are currently hard to predict.
ChatGPT has captured all the attention since its unexpected public release by Open AI late last year. It is a harbinger of the coming AI revolution. Even ChatGPT’s developers were surprised by the timing. They were given just 13 days’ notice to prepare the chatbot for release by their bosses, according to the New York Times. The goal of Open AI CEO, Sam Altman, was to be the first. It was a huge gambol. It turned the public into beta testers. But it has paid off. It has become the fastest growing app of all time. Open AI will soon launch a paid version, ChatGPT Plus.
It is still not mature. “ChatGPT is a horrible product.” That’s what Sam Altman CEO has just told the New York Times Hard Fork podcast. Behind the self-deprecation is a serious point. “Its cool for sure, but no-one would say this is a well-integrated product – yet,” Altman added.
Initial sceptics have certainly been won over. “I’ve changed my mind.” That’s what Azeem Azhar, an influential voice on AI has told 675,000 blog readers on 6 February. He explained how he has designed a sophisticated board game with complex rules in a dialogue with ChatGPT. He wrote “I want to say that this is likely game-changing. I don’t have other tools at my disposal that have helped me with problems like this (and the many others I have tested this chatbot on.)”
Azeem warns us not get too carried away. ChatGPT sometimes produces ‘hallucinations;’ seemingly plausible answers that are totally untrue. I searched my name and the book I co-authored with Nik Gowing, Thinking the Unthinkable. ChatGPT came back: “I am familiar with Chris Langdon, who is a co-author of the book “Thinking the Unthinkable.”….He is known for his work on the potential of new plant breeding techniques, such as CRISPR, to address global food security challenges. His book, “Thinking the Unthinkable,” co-authored with Mark Lynas, explores the implications of using these new techniques to improve food production and sustainability.”
Mark Lynas is a distinguished environmental writer. But Mark and I have never met, let alone written a book together. Not a single fact is correct – though it looks as if it could be.
At least, Google does have the right results about Nik and my book. But it is Google which has been the first casualty of the chatbot war. It has been working on large language models for many years. It decided to rush out its own chatbot, Bard on February 6. It was a fiasco. Bard got some basic facts wrong. Shares in Alphabet, Google’s parent company fell fast – 7% in a week.
To add to Google’s discomfort, Microsoft’s release the next day of ‘new Bing’, the AI-enabled version of its search engine has been a success. “This new Bing will make Google come out and dance, and I want people to know that we made them dance,” boasted Satya Nadella, Microsoft CEO.
Google’s rush to unveil Bard launch has certainly damaged Google’s reputation in the short term. There are cruel jokes that Google has just put the ‘AI’ in ‘FAIL.’ It has also put Google’s vaunted search engine model up for challenge. “I wouldn’t want to use Google anymore,” says Sam Altman. He is not exactly the most objective observer, given his GPT language model powers the new Bing.
Google has huge AI expertise its own large language model LaMDA and vast computer processing capability, so it is likely to catch up in time. There are questions about Google’s culture of innovation that it must address urgently.
These chatbot wars and talk of arms races are fun for the IT geeks. What does it actually mean for users?
There is a new reality. Chatbots are swiftly becoming alternatives to search engines. With intuitive natural language searching, and dialogues rather than keywords, the user experience will transformative. Though we still do need to keep our ‘BS detectors’ on full alert!
Try out the new Bing - when you can. It has a waitlist.
There are new search engines such as neeva.com. It is ad-free and AI enabled.
Soon there will be Claude, a generative AI chatbot to be launched by Anthropic, a start up by former Open AI engineers who want to go safety-first.
Over coming months, the way users and organisations measure their effectiveness on the web is going to change radically.
The rise of chatbots will likely transform the economics of search.
Google and Microsoft will try and come up with new way to rake in the cash. Ads will appear alongside their chatbot searches, The Economist predicts.
Savvy web teams are already preparing for a new world of Google search analytics from July. They will no longer be able to give their clients and bosses a neat set of Google analytics to evaluate the success of their internet marketing.
A big change will happen on July 1 2023. Google says: “Standard Universal Analytics properties will stop processing new hits. If you still rely on Universal Analytics, we recommend that you prepare to use Google Analytics 4.”
Google Analytics 4 uses new algorithms. That mean that the analytics of web-marketing effectiveness that organisations have been relying in recent years on will no longer be available.
Web developers we speak to are trying to figure out what this means for them and their organisations.
All organisations should do the same. And super-fast. Because that is the rate at which change is taking place.
Read Chris Langdon’s article on: Are we ready for the coming AI revolution?