Posts about expertise

The next net: expertise

The net is yet young and needs to learn from its present failures to build a better infrastructure not just for speaking but also for listening and finding that which is worth listening to, from experts and people with authority, intelligence, education, experience, erudition, good taste, and good sense.

Here is a preview of one nascent example of such a system built on expertise from Samir Arora, the former CEO of Glam. Samir and I bonded a dozen years ago over the power of networks, when he came to my office to show me one history’s ugliest Powerpoint slides, illustrating how open networks of independent blogs at Glam bested closed systems of owned content in a media company. I had been beating the same drum. Glam later imploded for many reasons, among them the change in the ad market with programmatic networks and investor politics.

Next Arora quietly set to work on Sage, a network of experts. It’s not public yet — he and his team plan to open it up in the first or second quarter — and it’s a complex undertaking whose full implications I won’t fully understand until we see how people use it. But as before, I think Arora is onto an important insight.

He started with a manageable arena: travel and food — that is, expertise about places. That topic made it easier to connect what someone says with a specific destination, hotel, or restaurant; to compare what others said about these entities; and to judge whether someone was actually there and spoke from experience as a test of credibility and accuracy.

I’ve begged Arora to also tackle news but watching him grapple with expertise — at the same time I’ve watched the platforms struggle with quality in news — it becomes apparent that every sector of knowledge will need its own methodology. Sports can probably operate similarly to travel. Health and science will depend on accredited institutions (i.e., people with degrees). Culture will be sensitive to, well, cultural perspective. International affairs will require linguistic abilities. Politics and expertise is probably oxymoronic.

Sage began with 100 manually curated experts in travel. Asking the experts to in turn recommend experts and analyzing their connections yielded a network of 1,000 experts with opinions about 10,000 places. Then AI took that learning-set to scale the system and find 250k experts and influencers with their judgments about 5 million places. 250k is a lot of sources, but it’s not 1.74 billion, which is how many web sites there are. It’s a manageable set from which to judge and rank quality.

I probably would have stopped there and released this service to the public as a new travel and restaurant search engine built on a next generation Google’s page rank that could identify expertise and control for quality and gaming. Or I’d license the technology to social platforms, as Twitter’s Jack Dorsey has been talking about finding ways to present expertise in topics and Facebook is in constant search for ways to improve its ranking.

But Arora did not. His reflex is to create tools for creators, going back to his cofounding of NetObjects in 1995, which built tools to build web sites. At Glam, he bought Ning, a tool that let any organization build its own social network.

So, at Sage, Arora is creating a suite of tools that enable an expert who has been vetted in the system to interact with users in a number of ways. The system starts by finding the content one has already shared on the web — on media sites, on Instagram or Pinterest, on YouTube, on blogs, in books, wherever — allowing that person to claim a profile and organize and present the content they’ve made. It enables them to create your own channels — e.g., best sushi — and their own lists within it — best sushi in L.A. — and their own reviews there. They can create content on Sage or link to content off Sage. They can interact with their users in Q&As and chats on Sage.

Because Arora is also reflexively social, he has built tools for each expert to invite in others in a highly moderated system. The creator can ask in contributors who may create content and curators who may make lists. Sage is also going to offer closed networks among the experts themselves so they will have a quality environment for private discussions, a Slack for experts. So creators can interact with the people they invite in, with a larger circle of experts, with the public in a closed system for members and subscribers, or with the public at large.

And because the real challenge here is to support creativity and expertise, Sage is building a number of monetization models. You can link to your content elsewhere on the web with whatever business models are in force there. You can offer content for free. You can set up a subscription model with previews (the meter) and one-time purchase (an article or a book). You can sell access to an event: an online Q&A, an individual consulting session, an in-person appearance, and so on. And you can sell physical products — a cookbook, a box of hot sauces — directly or via affiliate arrangements. Plus you can accept donations at various tiers as a kind of purchase. Note that Sage will begin without advertising.

I hope this platform could place where newly independent journalists covering certain topics could build an online presence and make money from it.

All this is mobile-first. Experts can build content within Sage, on the open web, and — if they have sufficient followers — in their own apps. The current iteration is built for iOS with Android and web sites in development. (Since I live la vida Google in Android, I haven’t been able to dig into it as much as I’d like.)

Users will discover content on Sage via search on topics or by links to experts’ channels there.

After starting with restaurants and travel, Sage is expanding into culture — reviews of books and movies. Next comes lifestyle, which can include health. News, I fear, will be harder.

So what is expertise? The answer in old media and legacy institutions was whatever they decided and whomever they hired. In a more open, networked world, there will be many answers, yours and mine: I will rely on one person’s taste in restaurants, you another. The problem with this — as, indeed, we see in news and political views today — is that this extreme relativity leads to epistemological warfare, especially when institutions old (journalism) and new (platforms) are so allergic to making judgments and everyone’s opinions are considered equal. I am not looking for gatekeepers to return to decide once for all. Neither do I want a world in which we are all our own experts in everything and thus nothing. Someone will have to draw lines somewhere separating the knowledgeable from the ignorant, evidence from fiction, experience and education from imagination and uninformed opinion.

Will Sage be any good at this task? We can’t know until it starts and we judge its judgments. But Arora gave me one anecdote as a preview: About 18 months ago, he said, Sage’s systems sent up an alert about a sudden decline in the quality and consistency of reviews from a well-known travel brand. Staff investigated and, sure enough, they found that the brand had fired all its critics and relied on user-driven reviews from an online supplier.

This is not to say that users cannot be experts. As Dan Gillmor famously said in the halcyon early days of blogs and online interaction: “My readers know more than I do.” In aggregate and in specific cases, he’s right. I will take that attitude over that of an anonymous journalist quoted in a paper I just read about foundations requiring the newsrooms they now help support to engage with the public:

The people are not as knowing about a story as I am. They haven’t researched the topic. They haven’t talked to a lot of people outside of social circles. I read legal briefs or other places’ journalism. I don’t think people do that. It can become infuriating when my bosses or Columbia Journalism Review or Jeff Jarvis tells me I’m missing an opportunity by not letting people tell me what to do. I get the idea, you know, but most people are ignorant or can’t be expected to know as much as I do. It’s not their job to look into something. They aren’t journalists.

No. Expertise will be collaborative and additive. That is the lesson we in journalism never learned from the academe and science. Reporters are too much in the habit of anointing the one expert they find to issue the final word on whatever subject they’re covering: “Wine will kill us!” says the expert. “Wine will save us!” says the expert. As opposed to: “Here’s what we know and don’t know about wine and health and here is the evidence these experts use to test their hypotheses.”

What we will need in the next phase of the net is the means to help us find people with the qualifications, experience, education, and means to make judgments and answer questions, giving us the evidence they use to reach their conclusions so we can judge them. Artificial intelligence will not do this; neither will it replace expertise. The question is whether it can help us sift through the magnificent abundance of voice and view to which we now have access to help us decide who knows what the fuck they’re talking about.

(You can apply to join Sage for Experts here.)