2016 wasn’t the year of the expert – quite the opposite when you think about Brexit and Trump, and how people had enough of experts. 2017 is a bit like a big hangover, scientists still have a headache about what had happened and how to respond. So, what has gone wrong – and why does it matter to open science?I argue that we’re living in an unbalanced transitional period: The internet has helped to transform our communications systems, significant parts of the economy, and even politics. But science is still grappling with these changes, it cannot keep up with the new speed in politics, economics, and communications. It therefore struggles to maintain one of its key functions: Informing the public discourse by injecting robust evidence into matters which are relevant to public life. To achieve this, science must start to operate at web-scale – and I believe Open Science is the key to get there. When fundamental elements of societies change, unstable periods of adaption can follow. The 1920s and 1930s are a strong example: After the first world war, the introduction of universal suffrage in many countries required the sudden integration of millions of new voters into the political process and culture. In this disruptive period, populism and even fascism flourished. In the rebalancing after 1945, science was given an important role: Western parliaments expanded their committee systems, increasingly inviting scientists and subject matter experts to give evidence – and thereby greatly impact evidence-driven politics.Today, we’re at a similar crossroads. The internet, through social media, now allows people to engage more intensely in the public discourse than ever before. It has also facilitated a far more diverse media landscape. Generally, this change has the potential to improve the quality of our democracies: More diverse, responsive, and quicker public discourses should be a good thing. But they have also given space to bogus claims, fake news, and populist politics in a range of areas, including climate change, health, economics, religion, biology, and many more. Most likely, these views will continue to exist – but the question is how we counter them.That should be role of science. Ironically, today’s research systems are more advanced than ever, with better methods, tools, and data to produce superior research outputs. Where we struggle is the speed, efficiency, and agility of how we produce scientific evidence. As the public discourse steams ahead, fuelled by web-scale communications and media systems, being slow has a price: It compromises the ability of scientists to inject evidence into public discussions at the best time. No doubt, our scientific insights are more robust than ever, credible media outlets will still report about them, and scientists are still invited to tell parliaments and governments on what to do. But until all this happens, tweets might already have taken the public discourse in a very different direction. Therefore, we need Open Science to upgrade science to operate at web-scale, so that it becomes more agile, efficient, and speedy. For this, we need to remove bottlenecks in how we produce, deliver, and preserve a diverse set of scientific outputs. Ultimately this will make their production more transparent, increase their availability, and improve their utilisation. Research cannot have impact when papers remain in peer-review for 6+ months, if publications are locked away behind paywalls, if data isn’t available due to unclear anonymization requirements, or if other research artefacts (such as software) aren’t shared because repositories are not interoperable. For science, the problem is that the status-quo consumes resources which it does not have – if it wants to retain its influence in an environment where discussions evolve quicker and require injections of evidence in more flexible ways. Open Science allows us to pool our existing resources, exploiting economies of scale (e.g. when sharing datasets) and the long-tail of highly specialised research. For this we need to think about solutions that help to make (inter-)national science systems more interoperable and that create connections between different research communities. The great advantage is that Open Science can build gradually on existing infrastructures and stakeholders. It can therefore reflect the fragmentation and diversity which is so important to the progress and rigour of the scientific community. But at the same time, it can help to remove the administrative burdens, duplications, and bulkiness that currently comes with working in science. Open Science is therefore an agenda to modernise and relieve science – and help scientists do what they want to do: Drive progress and do good for society.