True innovation in search is overdue, but does a system that returns results in the form of an instant webpage offer a glimpse of a possible future? Not YOUR webpage, however.
As the slow divorce of Google and publishers gathers steam, on grounds mostly akin to unreasonable behaviour, the simple and powerful utility that search has provided to us as a route to audience must not be forgotten or disregarded.
Yes, it seems odd to state something so obvious. But in publishing's current economically embattled position, reliance on the current state of search under the current search leader is now a reliance on an unstable and unrewarding traffic source, particularly given the expected move into full production of Search Generative Experience - in which there appears to be little good news for publishers.
Despite a long history together, publishers now openly talk of Google as the enemy, not a friend. Google - well, it's the elephant, they don't seem to be thinking too much about the mosquito let alone how it might help contribute to its next meal.
It's a pity that such an adversarial mindset exists. Those of us lucky enough to experience the first generation of publicly available search engines haven't forgotten the moment we started to comprehend just how much information such systems could search, and what that might mean. It felt like a form of knowledge freedom, or a promise of it. Most of us wouldn't have expected to end up where we are now.
There have been several candidates for a different type of search experience over recent years. The default privacy settings on search rivals such as DuckDuckGo have proven popular with those seeking to isolate themselves from the worst of tracking (or the best, depending on your point of view), for example. Browsers Brave, Firefox, or even Tor, have similar approaches.
It's still the case that an actual search breakthrough of the magnitude of the original PageRank has yet to see the light of day, even if I'm optimistic enough to believe such a thing exists in a data research lab somewhere if only some geniuses can find the right way to use it.
Being that search still holds a place in our hearts, if not our publishing wallets, then looking for solutions in the presentation and attribution of search results using AI seems one promising route.
There's been a buzz around The Browser Company for a while. It actually got my attention by virtue of being founded in New York, not Silicon Valley, appealing to my notion that different thinking often comes from different locations. Currently, their AI-enabled Arc Search application (still based on Chromium ) is gaining some momentum, with promising results (the irony of using "promising results" when talking about search is not lost on me).
The interesting aspect of Arc Search is that it effectively returns a query webpage. In essence, The Browser Company is attempting to unite various aspects of information retrieval - the browser, search engine, AI assistant, and website.
So, a user could ask for the latest Taylor Swift updates and a page would be constructed and returned with an AI-generated recent summary of the world of Swift (that we all apparently now live in), various links, more information, and a "dive deeper" section with more links. A demo of Arc's new iOS mobile app and a summary of how it works for the user was published this week on X.
Certainly such an application may find a wider market. If the generated part is found to be reliable and the QA of link curation is good, it may have a rosy future. Those are quite significant "ifs" in the world of GenAI as we all know. Early days though.
The application's most significant play is in the way is it attempting to change the user's search experience. The actual webpage presentation aspect likely has appeal to the librarian in many of us: results returned in a format that feels concise and controlled. Sources for that webpage are still our main concern though as part of the publishing industry, and as Gael Cooper writing for CNET said: "As a book author and journalist, I was concerned about how Arc Search offers credit and clicks for those who did the research work in the first place. The six sources it used to get my answer were listed briefly as Arc Search did its thing, but they didn't seem to be clickable. I've reached out to the makers of Arc Search and will update this story if I receive more details about how it credits sources."
There's the rub, as they say.
It's also clear there's still a wider issue here, a permanent caveat for publishers that any tech business such as The Browser Company is in existential peril of rapid onset enshittification - the American Dialect Society's 2023 word of the year, no less - starting out as good guys helping two existing parties (here, publishers and audiences) meet effectively in the middle and making the world a better place, before rapidly transforming into commercial villains by relentlessly extracting increasing rewards from both of the sides it serves.
This means any publisher must treat with extreme wariness any interloper trying to make their tech business big by using publisher content.
There is a future for search, and it is developing gradually. It's likely the search market will fragment more, and even segment, so different types of users get different types of search. An application such as that which The Browser Company has developed is an interesting step in a different direction.