🧙🏼‍♂️ OpenAI reveals SearchGPT

Howdy, wizards.

For months, there’s been speculation about what OpenAI had brewing in terms of a search engine. Yesterday they revealed their working to integrate search engine features in ChatGPT, and they’re testing this using a prototype called SearchGPT.

I’ve put together a comparison on the emerging generative search engines for you below.

Dario’s Picks

The most important news stories in AI this week

1. OpenAI reveals SearchGPT. It's a search engine that gives answers generated with AI, citing relevant sources. From the demo, the difference between SearchGPT and simply using ChatGPT with web browsing enabled, seems to be that the user interface is more tailored to search and speed. However, this is just a temporary prototype that is released to a small group of testers at the moment. The vision here is to use it as testing ground for bringing more powerful search capabilities back into ChatGPT. See how it looks below.

   ‎

   ‎ Why it matters‎ ‎ Credible rumours already made it clear OpenAI was working on a search product, we just didn't know which direction it would take. That seems more clear with this announcement. For now, I'd say it looks like a version of perplexity with slightly better layout and clearer attribution to sources paired with the answer (Perplexity uses almost invisible footnotes, while OpenAI's demo shows labelled blue links). Keep in mind, this isn’t a full product, but rather a playground for them to test features – nothing the majority of us will be able to use yet.

   ‎

2. Bing launches generative search. Similar to SearchGPT, it's currently being tested on a small number of users. Search results show an AI-generated page with a layout including a summary (with text, images, tables and more), related sections and sources; everything happens alongside traditional search results. See how it looks below.

   ‎

   ‎ Why it matters‎ ‎ Bing's search experience reminds me of Google AI overviews combined with Perplexity Pages, but with more prominent credits/references to websites in the results. Actually looks pretty nice. According to Bing, "early data indicates that this experience maintains the number of clicks to websites and supports a healthy web ecosystem"; this is huge if it holds true on a bigger scale, as many publishers are dependent on search traffic for their livelihood.

This issue is brought to you by

Grow Your Business with HubSpot's Free CRM

HubSpot offers an intuitive customer relationship management platform tailored for small businesses. Manage leads, track sales performance, and understand your customers with ease. Best of all, it’s completely free, with no limits on users or data, allowing you to store and manage up to 1,000,000 contacts.

Get HubSpot CRM free today

Behind the news

A comparative look at AI search engines

To help you (and myself) understand a bit better how the emerging GenAI search products differ from each other, I’ve put together a little comparison below. Note that this was just a simple test I did based on the screenshots from the demos.

OpenAI’s SearchGPT

Search query: “music festivals in boone north carolina in August”

Versus Google: SearchGPT shows more useful results for this query than traditional Google results, imho.

Versus Perplexity:

• SearchGPT has clearer in-content attribution (uses labelled blue links vs barely visible footnotes) and in-content images

• SearchGPT also gives you traditional search results in a sidebar

• SearchGPT bases each festival recommendation on one source, while Perplexity sometimes uses multiple sources

SearchGPT seems a step above Perplexity in the sense that it combines a more traditional search experience, and also gives clearer attribution to publishers. The former is helpful to users l in cases where you want to explore content on the topic directly, and the latter is crucial to maintain publisher revenues and supports high-quality content on the web.

Bing’s generative search

Search query: “how long can elephants live?”

Versus Google: For this query, Bing’s generative search seems only slightly better than a traditional Google search. It might be because the answer to this query (Elephants’ lifespan) isn’t changing much over time, and Google already has a well-adapted interface for that.

Versus Perplexity:

• Bing offers traditional search results alongside the GenAI answer, both in-content and in a sidebar.

• Bing has much better layout in its GenAI answer: the answer to the query is the heading, a relevant video appears alongside the answer, a neatly formatted table differentiating lifespan by species appear below.

• Bing has the most clear attribution to publishers with labelled links including the website’s favicon (logo)

Bing seems a good step above Perplexity, both in how helpful the result is and how well it attributes sources.

‎ Why it matters‎ ‎ The future of the web is being laid right now. If Google, Microsoft and the AI players releasing search engine experiences don’t find an effective way to compensate big and small publishers, the incentive to put out original, high-quality content on the web could vanish – nobody wants that, nor users for whom the web would be flooded with regurgitated AI non-sense or the AI companies who rely on quality content to train their models.

I’d say Microsoft Bing shows the most promising search layout with AI right now, and might be in a unique position to develop something great. The company can build upon vast data on search behaviour from its existing search engine and doesn’t have to worry about shattering a $200 billion ad business while experimenting like Google does.

Google has a more difficult task due to its need to incentives to maintain an extremely profitable advertising business – a core part of Google’s business – with creating useful AI features. Microsoft on the other hand, has more freedom to experiment in that most of its revenues come from other sources.

The wizard’s favourite AI newsletters

what i’m reading right now

TLDR 💨 - essential tech news in 5 mins. Read by 1.2 million people

simple.ai - 🕵🏻‍♂️ - deep-dives on AI by Hubspot’s co-founder

Bagel Bots 🥯 - best hands-on tips & tricks

The Neuron 😸 - easy weekday read on AI’s latest developments

Affiliate disclosure: To cover the cost of my email software and the time I spend writing this newsletter, I sometimes link to products and other newsletters. Please assume these are affiliate links. If you choose to subscribe to a newsletter or buy a product through any of my links then THANK YOU – it will make it possible for me to continue to do this.