Artifact, the news discovery app launched a year ago by the founders of Instagram, has just announced it is shutting down, saying “the market opportunity isn’t big enough to warrant continued investment in this way.”
As pointed out in TechCrunch, Artifact “hadn’t quite defined what it wanted to be”, iterating from its initial personalized news discovery platform to include conversations about news articles and the essence of a social network.
Given my lifelong focus on information discovery I was excited by the launch of Artifact and was an early user. I found it somewhat useful in surfacing interesting articles, but not significantly better than other similar apps.
News discovery still sucks
Having closely followed the information discovery space for over two decades, I continue to be amazed at how poor our tools are.
In my recent book Thriving on Overload I wrote:
Problems are the spark for invention. It is safe to say that we have a deep problem in our ability to find the most relevant information and avoid irrelevant or misleading content. We need better tools than we have today.
Part of the reason has been in the creative destruction that has been emblematic of the Internet economy, leading to the untimely death of useful platforms such as Delicious and Friendfeed. The rapidly mutating economics of the news industry has also played a role, with a proliferation of aspirant news aggregators, each jostling for a share of advertising revenue, perverting many efforts. The potential value of personalized news feeds has been undermined by (usually merited) distrust of the companies that apply behavioral data to peddling advertisements rather than creating value for their users.
There remains a massive opportunity to provide better services to help us filter and find what is most useful to us. It is highly encouraging to see the increasing pool of talent, inspiration, and capital being dedicated to help us discover the most useful information, and distinguish between information that is correct, questionable, or plain wrong. I hold strong hopes that in the current mix of contenders in this space some truly valuable services will emerge and prosper.
The recent leap in AI capabilities provides far more powerful tools for information discovery, personalization, quality assessment, and presentation.
Yet I am still not seeing the tools or apps that can provide us with any semblance of what we need to know each day.
Human curation prevails
This is underlined by the surge in newsletters as people strive to keep apace with developments in AI, technology, and society, combined with insightful commentary.
While many newsletters are AI-assisted, almost all are at their heart human-curated selections. This still leaves individuals with the task of finding and selecting the set of newsletters that best matches their unique set of interests.
Scope of the opportunity
The size of the opportunity for news discovery has been evident since the mid-1990s, when it was clear that the Internet would become a far better place for information discovery than traditional media.
Attention has become the most valuable resource. Being the portal for daily (or usually more frequent) updates from many users merits a massive valuation.
Among the attention giants, Google News and Yahoo News were in the vanguard for single source news discovery, with Apple News coming later. They are all reasonable products, but for anyone who is interested in going beyond the surface they are pretty poor, even in their most personalized forms.
Much news discovery happens on social networks such as Facebook, Twitter, and TikTok, making this a significant aspect of their enterprise value. However by themselves they are limited as tools, at best in the case of Twitter requiring a lot of work to surface the most interesting news.
Challenges for information discovery
Why has no-one has cracked what would seem to be a tractable problem decades after it was clearly identified as a major business opportunity?
I point to some of the challenges in the excerpt from my book. Other include:
- News personalization requires extensive data. Privacy concerns and regulations can limit user information capture.
- Monetization can be hard, as Artifact discovered. Discovery is built into the major platforms’ offerings, but they don’t necessarily have incentive to provide what is most useful to the user.
- Sticky hebaviour. It is hard to attract users to new services when information habits are often deeply engrained.
- People don’t know what they want. Relatively few people have clarified what information is useful to them (a core them to Thriving on Overload). Mapping their current information habits does not necessarily lead to better discovery.
There are other challenges. Yet these are not overwhelming relative to the opportunity.
Who will take us higher?
Who out there is creating something which is a level up from what we have now?
We have the technology, particularly with the advent of generative AI, there are business models to make this work, and smart people are intensely interested in this issue.
I don’t think I’m asking for too much to have quality, personalized news and information discovery in 2024.
[Postscript: For some commentary by smart people on this see the LinkedIn post.]