Stephen Webster 00:04
But I just want to welcome you, those of you who are external to Imperial.
Of course, those of you who are at Imperial, welcome too.
It’s not often we can get together as colleagues to discuss these kinds of things.
That’s me kicking off the opening session of the Failure in Science conference.
I’m Dr Stephen Webster, director of the Good Science Project at Imperial College London, a programme that examines what we like about research culture and what we don’t.
And of course, being a university, that means conferences.
And this one, Failure in Science, has really got a lot to do with Magdalena Skipper sitting here because she said to me this time last year, ”Should we have a conference on failure in science?”
And I just thought, Imperial College, Nature magazine, a conference about failure. You know, this is really too good to be true. Let’s work on this.
So, and so Magdalena Skipper, (she’s editor-in-chief of the journal Nature), and I got to work.
And in April 2026 at Imperial, we welcomed 150 delegates, all of them eager to talk about something not much talked about: failure in science.
We spent the day focused on stuff going wrong, mistakes being made, opportunities being missed, and good things just being neglected.
In this series of six podcasts, using the conference as a springboard, we’re exploring different aspects of failure.
That means much more than simply looking at failure as something personal, a spot of bad luck coming our way.
When it comes to science, is there institutional failure too? For example, should universities frame failure not as a problem, but as part of a creative process, as part of good science? How can institutions look more sympathetically at failure and see failure as a necessity when quick progress is so often the bottom line.
And we’ll also look at the impact of science more widely. What can go wrong when scientific knowledge leaves the lab and tries to make a difference in society?
But first, in episode one, over to Magdalena Skipper.
Magdalena Skipper 02:41
First of all, actually, I want to start by thanking you and Imperial for following up on what I honestly did not think you were going to follow up on.
Because failure is a taboo topic.
Even though we know that failure is part of success, failure is one of the best teachers that we have, as long as we learn from those failures and don’t repeat those mistakes.
I’m Magdalena Skipper. I am editor-in-chief of Nature.
At the conference, I talked about the fact that failure is often a taboo. We don’t talk about it. We know that in science, failure is an everyday occurrence, and of course, failure itself is a very large umbrella term.
Most experiments fail. We learn from that failure, a great deal. Sometimes our plans in our career fail. We learn from that as well. We have to back up and pivot in another direction.
Considering how useful those lessons are, how contextualized they are, I find it surprising that we don’t talk more about failure, that we’re not more honest about the role that failure plays.
But we are petrified of it.
If you failed, you’re a loser, and that’s why nobody wants to talk about it.
Hence, it’s a taboo.
Failure is part and parcel of our lives, be it professional or personal. We learn through failure. There isn’t a single individual who hasn’t failed at something.
The really crucial aspect of it is how much we learn from it. We learn it’s a matter of building resilience, but it’s a matter also of delving deep into the topic, into the situation in which we have failed, and understanding it much, much more deeply in a much more nuanced way.
So, on the one hand, it’s a natural process. There isn’t a person or a system that hasn’t failed at some point. And on the other hand, it’s such a useful teacher.
The conference was part of a wider project called the Good Science Project, and in the end, the day focused on a number of levels.
One was an individual level. How individuals fail themselves and how they feel about it, how they recover from it, what they learn from it.
We talked about what failing means at an institutional level, or an organizational level.
And of course, since science or research more broadly doesn’t just happen in an academic setting, we also talk about how failure is dealt with in other settings, in the private sector, in in society more broadly.
Much later on in my professional career, I made a decision to step away from scientific, active scientific research and become a scientific editor.
I did that after a PhD and after a postdoc. And now, for me, looking back over the years, I think that was, without a doubt, the best professional decision I ever made. But to some who are wedded to the idea that the only true path for a researcher is to continue in active research, that was a sign of my failure.
I failed to be a successful academic researcher.
And it’s really interesting because that one act, that one action that I took, for me, was a manifestation of success or successful decision, and a manifestation of failure for others.
So I have been referred to as a failed scientist. And of course I’m a failed many other things.
Once upon a time, I wanted to be a firefighter. I’m also a failed firefighter. That never came to pass.
I think the most important thing is that I don’t fail all the time at what I’m actually doing right now.
And perhaps the next thing to add is, in general, in failure, if you do fail, learning from that experience so you move on, and you move on as a richer individual, a richer professional, is really important.
Think about the following example...
If you are constantly petrified of failing, you are going to play safe all the time. You’re not going to take risks. You’re not going to place big bets, if you like.
Now, in science, that translates into addressing big questions, formulating risky hypotheses, deciding to dedicate funding to a risky, perhaps long-term project that is not fail-safe, is not something that, broadly speaking, we already understand, and you colouring in the boxes or filling in the blanks.
No, if you are not afraid of failure, you are more likely to take bigger risks. And of course, when you take bigger risks, yes, there is a risk of failure.
But there is a risk of there’s a possibility of of great of great gain. Now, our general aversion towards failure means that in our funding systems, we tend to we tend to prefer to fund projects that have quite a lot of foundation, that is safe already.
Many scientists have the experience when they apply for grants, that a large proportion of what it is they’re proposing in the grant they already know is going to work because they have demonstrated it at some smaller level to a smaller degree.
So the question then becomes: is our funding system really designed to push the boundaries of knowledge and make us take those risks and discover something completely unknown, take us in a new direction, maybe open a new avenue, maybe even create a new discipline, or uncover a new approach to probing some questions?
Or is it more designed to walk, take very small steps in some direction?
It’s fascinating to consider the different attitudes that different sectors, different industries have towards failure.
But my understanding is it’s a practice that is relatively widespread in technology companies, especially in Silicon Valley.
And the practice is as follows. In order not to stifle and prevent that risk-taking attitude, which may lead to failure, those companies actually reward their employees financially for failing.
But if they fail fast, imagine without this incentive, you have a situation where only the employees-and this is in the research setting within Google X, only the employees that come up with quote unquote ”positive results”, something that can be taken forward, are rewarded.
But those positive results may be less impactful, or less influential, or have fewer implications if all the employees are thinking: Oh, I will only get rewarded if I succeed. I therefore must not fail at all cost.
That really influences your risk taking.
Now, of course, from a company’s perspective, there’s no point investing employees’ time or money in long-term pursuit of a project that fails. So what is being rewarded is fast failure. That is crucial.
That is failing fast. To me, it’s a really interesting way of incentivizing risk-taking.
Basically, ultimately, of course, everyone wants to succeed, but the road to that success often leads through failure, and maybe there isn’t another way. Maybe you have to go through failure in order to get to that success.
And contrasting that to an academic setting, where often a lot of time and money is invested in long-term projects, then there should have been a stop, no go decision made earlier on, but perhaps I’m sure there are a number of reasons why a similar approach is not adopted in academic setting.
But one of them I would suggest is that fear of failure? Once you have a project, it’s your project, it’s your baby. You’ve created it. You need to continue proceeding with it.
At the conference I mentioned one specific example of how failure can really energize and motivate researchers.
So the specific example, it’s an example of a paper, a scientific paper, describes a metascience study, so a science of science, where the researchers obtain data from a funder, a US funder, looking at a population of scientists around the cutoff bar for funding.
So they looked at a small, small sliver of this population just above that cutoff. So individuals who just scraped in and got the funding. And then those individuals who just didn’t make it, who just didn’t get the funding. They followed those researchers through time.
And the really interesting finding is that those who just failed to get the funding did better in terms of their career, establishing their career, and contribution to science, than those who just got the funding.
Now that was an observation in the paper. We can speculate why there might be, but I myself find it particularly compelling to think that it’s almost a demonstration of a saying: What doesn’t kill you make you makes you stronger.
That if you’re really convinced that you have a great idea, that you are committed to pursuing a scientific career, and you failed in gaining this support, financial support, then you will find another way.
And that induces that inventiveness and creates a greater motivation for you to persevere, because you are sure that you need to succeed.
So here, failure gives you a boost, if you like, is an additional motivator.
We all know, we have discussed already, that science progresses through self-correction, through a series of failures. These are learning opportunities.
Every graduate student, every scientist, working scientist in the lab, whether it’s the wet lab, or computational lab, or in the field, knows that a large proportion of experiments are going to fail, and that setup and expectations need to be adjusted.
Maybe the question needs to be modified. Maybe parameters need to be changed. Again, that learning process, that comes out of failure.
But we never talk about it. We never talk about that process, that iteration, that trial and error, (it’s not random. It’s intentional), but it’s there. We never talk about it when we talk about science to the general public.
Sometimes the rhetoric is more reminiscent of religious belief than actually a scientific fact that has been arrived at through that iterative process, stops and starts, and failures, and consolidation of ideas, and changing the hypothesis, etc. etc.
It’s not that the scientists don’t know what they’re doing, that they’re not following the scientific method, they are, but they’re not communicating that to the public.
I think the best example of that was during the pandemic, the COVID pandemic, when the general public, I have a sense the general public got a little bit confused and tired of scientists advising policymakers and then policymakers making decisions that appeared to flip flop.
You know, "now do this, now wear face masks, then maybe don’t wear face mask, maybe isolate, maybe don’t isolate.”
There were changes of recommendation and position. Now, to somebody who understands the scientific process, this is to be expected. We are in a completely new situation.
Nobody who’s alive today, who does research on the pandemic, has ever experienced a pandemic of this scale. So of course we’re learning as we go along.
It’s natural that it’s going to lead to failures, which then can be corrected.
And again, this is so crucial to emphasize it.
We learn from these failures, and the next decision, the next recommendation, is going to be better informed, better underpinned by facts.
And so taking a step back from this example, communicating that scientific process, which is all about failing, learning from failure, moving on, maybe failing again, taking a step back, taking two steps forward, maybe another step back, and and so on.
I think by missing that opportunity, we miss the opportunity to make science and scientific process more trustworthy, to show that it really is worthy of trust. We talk about trust in science a great deal, rightly so.
We are worried about the fact that trust in science is dropping.
Perhaps by embracing failure, we can reverse that trend.
We can be more transparent, more honest, and therefore bring more people along.
It’s an interesting point to consider whether a conference on failure should itself be a success or failure.
I think whatever we discuss here will be important, will be success.
And I will just end with a quote that I came across by complete coincidence in a bit of a trashy novel that I was reading on holidays recently.
I was very surprised. The quote comes from Winston Churchill, and apparently he said: "Success is the ability to go from one failure to another with no loss of enthusiasm.”
And I actually think that summarizes many of the endeavours that you all engage in, whatever you do.