Looking back at how we approached AI news coverage when we launched versus now, several things have genuinely changed - not cosmetic adjustments, but real shifts in judgment about what matters and how to report it responsibly.
We became more skeptical of funding-round coverage specifically
Early on, we treated a funding announcement as newsworthy largely on its own terms. We've since become more selective, prioritizing what a round actually signals about the business over the raw dollar figure, since the latter alone tells readers less than it initially seemed to.
We slowed down on same-day product launch coverage
We used to prioritize being first to cover a product launch. We've shifted toward being accurate over being fast, often waiting a day or two to actually test claims rather than repeating a company's own description of what a new product does.
We got more careful about anonymous sourcing
Early coverage occasionally leaned on loosely-sourced claims more than we're comfortable with in retrospect. We've tightened our bar for what counts as sufficiently corroborated before publishing.
Why we're sharing this rather than just quietly updating our practices
We think readers evaluating any publication's coverage benefit from knowing that editorial standards evolve, and specifically how, rather than assuming a publication's practices are static from day one. These changes reflect things we got wrong often enough to notice and correct.