News stories glue portfolio managers and analysts to their screens. Each story feeds into a positive or negative bias. What if I could automate that? What if I could read everything as it comes out and sort it according to positive and negative news for a company? I could react systematically, quickly and across more stocks. I might even replace the role of the analyst.

Machine-readable news starts with exactly this economic intuition.

Disaffection with existing quantitative trading signals has brought it to investor attention. Advances in linguistic processing and the steady decline in computing costs have made it better and cheaper than before. And a raft of academic and industry papers have guided the way.

The first generation of machine-readable news providers aimed to do just that. They started with two objectives. First, figure out what companies the news is talking about. Second, bucket the news into positive and negative stories. But their approach was shaped by the idea that information is scarce – perhaps a vestige of the economic legacy of print journalism, closed electronic news networks, and the hot news doctrine.

The news-wires created captive providers of machine readable news. Chief among these is Thomson Reuters NewsScope, followed by Bloomberg and Dow Jones. These offer feeds based on the news generated by the parent company’s editorial production. As captives, they can advertise special access to news content, the editorial process, and other resources.

The strategy behind the first entrants, however, suggested that not everyone with a news-wire needed to develop their own platform. Instead, one could establish a strategic relationship with a technology provider. Ravenpack, for example, partnered with Dow Jones to provide its first generation machine-readable service. They have since launched a partnership with the AP to bring the news-wire to the financial community in machine-readable form.

The machine readable news feed was designed to be fast and straightforward. Early successes stemmed from low latency access to economic data from government agencies, such as the Department of Labor. Press-access to the “lock-up” favored news organizations, and providers found they could easily collect the figures, distribute the data quickly, and present it in a way clients could use it.

Machine readable versions of news-stories followed a similar approach. These started with the document and asked it questions. What companies is it about? Is the document positive or negative on those companies? What events were discussed in the article? These questions boiled the article down to its essentials, so it could be delivered and processed quickly – just like the economic data.

Information scarcity characterized the first generation of machine-readable news providers. It was a service from publishers, through which the financial community could get a machine-readable scoop. Even the name suggests a special relationship with a publisher – news comes from a publisher, so machine-readable news must also come from a publisher. Unsurprisingly, the first entrants were captive providers and strategic partnerships that sorted news articles according to positive and negative signals for companies. The problem was, information isn’t scarce.

Without question, the internet and the corresponding shift to digital dramatically increased the diffusion of information. The shift to digital also means privileged access to a wire-service provides little to no benefit over the manifold echo-chamber of the web. If it’s in a wire-service, it’s on the web, so it’s unnecessary to have special access to an editorial source.

Privileged access may also prove a hindrance. The flip-side of having a scoop is missing a scoop. Access to the machine-readable Dow Jones feed is great, but what about what the other wires are broadcasting? What about what other news or media sources are discussing? What about blogs and social media?

The first generation of machine-readable news providers invented a new way of consuming and organizing the news, but it’s just the first step. It was called machine-readable news for a reason. It was predicated on special access to the publisher and colored by the legacy of so many artifacts of the news industry that have exhausted their relevance – the hot news doctrine, print publication, closed networks.

While the first generation’s initial steps were revolutionary, the shift to digital has changed what’s possible and what’s necessary. The next generation requires access to a broader mix of sources, deeper insight into the underlying content, and a departure from merely taking the score of sentiment, positive or negative. And the shift will create an opportunity for new entrants to disrupt the old guard.