Help us write better implications
Our engine reads pharmaceutical news and writes what each item means competitively — what it changes, who it pressures, what follows. We can measure whether it retrieves the right evidence. We cannot tell whether the conclusions it draws are the ones that matter to you. Only you can.
So we have made the task as small as we honestly can. You will see 25 news items. For each one the engine has written two implications from the same evidence, taking different views of what the news means. One question per item: which one is better?
Nothing else is required. If one of them gets something wrong you can say so in a note, but only if you want to — the choice is the part we need. About 19 minutes in total.
Your choices are used for one purpose: to improve the implications this engine writes. They tell us which framing to favour, which evidence to lean on, and where our reasoning does not survive contact with someone who knows the field.
They are not an assessment of you, they are not shared outside the team building this engine, and they are not used for anything else. A disagreement is worth more to us than a compliment — if you think both options are wrong, say that.
Your answers are recorded here as you go, so you can close the tab and pick up where you left off. At the end you can download your own copy of everything you entered.