Six degrees of separation

My brother, David Silver, is panning out to be one of the significant players in the world of artificial intelligence. His PhD topic was applying reinforcement learning to the oriental strategy game of Go, and he has gone on to be the lead researcher on AlphaGo at Google DeepMind. That is the program that last year beat the world champion human player and became the best computer player of Go. More recently AlphaZero has taught itself to play Go from scratch (AlphaGo started by learning from thousands of top level human games) and has also taught itself to play chess and shogi, all to unprecedented levels of excellence. It has been very exciting following his progress, and going to the premier of the documentary film about AlphaGo (which is a lovely human drama, even if you don’t know or care much about the technology, so do give it a watch on netflix/prime/google play/itunes if you get the chance).

It is no surprise to me that David has gone on to find a niche that is intellectually impressive, as he has always been a pretty smart guy and done exceptionally well in education (though reassuringly he isn’t all that practical, makes the same silly mistakes as the rest of us, and has remained quite down to earth). I’ve always been glad to be the older sibling, as I think it would have been difficult to follow in his footsteps. As it was, I could be proud of my relative achievements before he came along and beat them all! He has always had a very analytical mind and enjoys solving logical puzzles. I guess I do too in some ways, but I’m much more interested in how people work than complicated mathematical calculations, and how we can reduce suffering and help people recover from trauma, rather than pushing the boundaries of technology. We’ve chosen quite different career directions, but I think we still have quite similar underlying values and ethics.

Although I’m proud of him, I’m not mentioning my brother’s achievements to show off (after all, I can take no credit for them) but because they’ve given me cause for reflection. Firstly, it would be easy to feel inadequate by comparison. After all, he is making headlines and working on the frontiers of technology, whilst I’m just a clinician running a tiny company and have made relatively little impact to date. It would be easy to be jealous of the financial security, publications and plaudits that he has got. He has made the news all around the world, and even has a wikipedia page! But I think I’d find that spotlight uncomfortable, and I suspect I’d find his job pretty stressful, as well as finding all the maths and computing pretty boring and unfulfilling. So whilst there is plenty to admire, I don’t really envy him and wouldn’t want to swap places.

Secondly, and perhaps more interestingly in terms of this blog, it has made me think about what my goals are. Making the best possible AI to play Go is quite a narrow and specific goal, and within that he selected a specific methodology with reinforcement learning, and he has focused on that for the past decade, before looking at what other applications the same system might have. Yet in that same time period I’ve been pulled in many different directions. I’ve been an NHS CAMHS clinician and service manager. I’ve been an at home mum. I’ve helped to found a parenting charity. I’ve set up and evaluated a project to improve outcomes for diabetes patients. I’ve bid for grants. I’ve tried to help recruit psychologists and improve clinical services within a children’s home company. I’ve undertaken specialist assessments of complex cases. I’ve been an expert witness to the family courts. I’ve delivered training. I’ve run a small therapy service. I’ve conducted research. I’ve tried to influence policy, and sat on committees. I’ve written a book about how to care for children affected by poor attachments and trauma. And I’ve developed outcome measures. Most of the time I’ve done several of these things in parallel. It is hard to keep so many plate spinning, and means I have not been able to invest my full energy in the things I most want to do. I’ve also had hesitations about investing in entrepreneurial ideas, because of guilt about saying no to other stuff, or fear that it won’t pay off  that have taken a really long time to shake off.

Greg McKeown says in his brilliant articles for Harvard Business Review about ‘essentialism’, that success can bring on demands that cause you to diversify, and ultimately reduce your focus on your primary goal and cause failure, and that is exactly what I’ve experienced. It reminded me of a reflective exercise I did as a trainee on a workshop about creative methods, where I made an amoeba shape out of clay to represent the pulls I felt in different directions. The amoeba was a resonant image for me as it can’t spread too thin without losing its depth at the centre, and it can’t travel in two directions at once. Finding the right direction of travel and resisting other pulls on my time is something I am still working on 20 years later! It has been a growth curve to learn what to say ‘no’ to so that the company does not become overloaded or incoherent*. There are also other forces that influence what a small business can deliver – we have to do work that we are passionate about, uniquely skilled to deliver and that there is a market for. There is no point offering services that nobody wants to buy, or that other people can provide better, or that you are not enthusiastic about, so we need to stick to things that we can deliver brilliantly and build a positive reputation for. However, with the breadth of clinical psychology there will always be multiple demands and opportunities, and it is necessary to find a focus so that we have a single defined goal** in order to attain the most success.

I’ve taken time to refine my goal from “applying clinical psychology to complicated children and families facing adversity” (which is actually quite a broad remit, and includes a wide range of neurodevelopmental, mental health, physical health and social aspects of adversity, being applied to all sorts of different people) to “applying clinical psychology knowledge to improving services for Looked After and adopted children” to “using outcome measurement tools developed through my knowledge of clinical psychology with placement providers and commissioners to improve outcomes for Looked After and adopted children”. Likewise, it has taken me time to clear space in my head and in my diary, and to be in good enough physical health to give it sufficient time and energy. But I am finally able to dedicate the majority of my working time to making people aware of BERRI, doing the statistical analysis to validate and norm it, and supporting/training those who subscribe to it. I have secured an honorary research fellowship at UCL and some data analyst support, and a trainee from Leicester is making it the subject of her doctoral research, so I very much hope that 2018 will be the year that we publish a validation of the measure and methodology, and can then roll it out more widely. I believe that is my best chance to make a difference in the world – to improve the standards of care for children living outside of their family of origin by encouraging universal psychological screening, regular outcome measurement, and the ability to identify and track needs over time.

Finally, my brother’s achievements have given me pause for thought because him working at Google has made me feel a sense of being somehow distantly connected to silicon valley, and all the technological and entrepreneurial activity that goes on there. Suddenly the people who founded Google, Facebook and Tesla/SpaceX are no longer as abstract as Hollywood actors or international politicians, but are now three steps away in a technology game of six degrees of Kevin Bacon. It makes the world feel a little smaller and making an impact seem more possible, when your kid brother is connected (however peripherally) to the technology giants who are changing the world.

Alongside this, in my ImpactHub coaching peer group several people have gone on to make successful social businesses that have rapidly scaled and made an impact on the world. Proversity for example, have expanded massively into the digital education space. Old Spike Roastery & Change Please have expanded their coffee businesses that employ homeless people, and School Space have scaled up a project they started at the age of 17 to help their school rent out its premises out of hours into a thriving business that has generated £350,000 of income for participating schools. Code Club have partnered with the Raspberry Pi Foundation to teach children in 10,000 clubs in 125 countries all around the world to program computers. And Party for the People have made a competitor for TicketMaster or SeeTickets where the fees go to a good cause, and have set up arts spaces in old factory buildings.

In this context, it seems possible to dream big, to think that an idea could become a reality that has an impact on the world. So whilst my main vocation remains to bring the process of regular outcome measurement to services for Looked After Children (and that is making some really positive steps at the moment), I’ve started to work out how to make my back-burner project a reality. This one is a proper entrepreneurial idea in the digital space and tied in a little to my previous blog topic of the issue of how the public understand the evidence for different kinds of interventions. I’m hoping I can develop a pilot and then seek some investment, so watch this space as I’ll report back how it goes.

In the meanwhile, I still want to make some changes in my personal life. I’m generally feeling quite upbeat about the future at the moment, and I’ve sorted out the issues I mentioned in a prior blog about disappointment. We’ve also pulled in payment for many of the outstanding invoices, and the business is the best organised it has ever been. But after reviewing how I spend my time and who I interact with the most, I have become much more aware of my various different networks, and to what degree I feel able to express myself authentically within them. I am being a bit more thoughtful about my networks, both in real life (where I want to make greater efforts to meet like-minded people locally) and online, where I need to spend less time (and I’ve recognised my shamefully judgemental feelings of rage and disgust when interacting with people using image filters, especially those that put puppy ears and noses onto reshaped faces with features made to conform with patriarchal standards of beauty). I have realised that I haven’t been choosing the company I keep well enough, so I am trying to connect more  with those who are positive influences on my life, and to pull away from people who are a drain on my emotional resources. I am also choosing to engage more with people in the social entrepreneurial space. As Jim Rohn is much cited as saying “you are the average of the five people you most associate with” and hanging out with inspiring people allows us to be more creative and entrepreneurial ourselves.

So hopefully 2018 will be the year where I make a success of BERRI, complete the validation research and get some publications out. I’d also like to get a pilot of my entrepreneurial idea up and running. And in my personal life I’d like to get back to the gym, to get the planning permissions sorted out for my house, and most importantly to make more real life social connections with people who share my values. If I’m only a few degrees of separation from people who have achieved all of these things, then maybe I can too.

 

*I wrote more about developing my business model and setting up a social enterprise in clinical psychology forum number 273 in Sept 2015

**or failing that, a primary goal, secondary goal and fall-back plan, in ranked order of preference (with an awareness than only exceptional polymaths like Elon Musk can achieve in more than one area at the same time).

The double bind of trying to do research as a clinician

In my first job after university I was an Assistant Psychologist on an applied research project evaluating the impact of staff training on the quality of care in old people’s homes. It wasn’t my natural client group, but I loved the fact that we were measuring whether all our fancy psychology ideas actually made a difference to people when applied in practice. My supervisor, Esme Moniz-Cook was an inspiration in this regard, and set a pattern I have aimed to maintain to this day: innovate, evaluate, disseminate. We wrote lots of papers, and each one worked through many iterations in which Esme would cut up a print out of my latest hopeful draft and stick it back together and/or annotate it in different coloured pens. I developed a ritual to overcome the frustration: I would look at the result, take a few deep breaths and then set it aside. Then I’d move the text around on screen to match the arrows and sellotape. Then set it aside. Then do a few of the suggested edits that felt too trivial to argue about and set it aside. Then I’d look at the reduced number of suggestions that was left and decide which were worth disputing and do the rest. Then I’d send it back to Esme and the process would repeat. It was a challenging process, but the papers were always better as a result, and meant I started training with a set of publications on my CV. My doctoral research felt like a piece of cake by comparison. I developed and evaluated a computer based training tool to help people (especially those on the autistic spectrum) learn to recognise facial expressions of emotions and predict how people would feel in different situations. I published a paper from it, of course.

As a newly qualified clinical psychologist in 2000, I was all fired up about continuing to do research. I put in a bid to two charitable funds to expand my doctoral research to a larger population and to follow up the effects after a period of time. It took me a long time to complete the proposal document and to get all the signatures, references and endorsements from both academic and clinical hosts. Both were declined. The first said the bid was so strong that it didn’t need autism specific funding, the latter said that we hadn’t accounted for poor uptake or drop-out because they didn’t know how keen the kids with ASD had been to have extra computer time. It felt like a Catch 22. In retrospect I should have researched the potential funders better; the former probably wanted research that would relate to their patented use of pig secretin in autistic children, they focused on medical interventions and were not interested in something as ‘soft’ as psychology. I was gutted. In the time I had spent preparing the two bids I could have made significant inroads into the study, but I hadn’t got off the starting blocks.

In my next job, I applied my learning to undertaking small bits of research myself by juggling my clinical time. I explored what it was that defined the young people who challenged multiple agencies. I evaluated how social workers in the adoption support team reacted to having the opportunity for consultations with a psychologist. I measured the Theory of Mind skills of Looked After Children and compared them to children with ASD and a control group. I assessed the cognitive ability and mental health of the children in residential care homes within the catchment. Most importantly, when asked to advise on how to improve adoptive matching, I undertook an audit of factors affecting adoptive outcomes across 116 families. This turned out to be the largest study of risk and resilience factors in adoptive matching ever undertaken in the UK.

I found the answers to my questions, and I learnt a lot of new things that other people need to know. But I had no funded time or academic support to disseminate the results. With the smaller projects, I let my Assistants write up the results for publication in professional newsletters and present them with me at conferences. But with the larger studies I wanted to do them justice, and my NHS role never gave me an opportunity to do so.

Once I left the NHS, I followed up my adoption audit 5 years after the initial data collection because I was determined not to let such interesting findings disappear into the ether. Then I made another bid for research funding, and by fluke or serendipity secured a £75,000 Shine Award from the Health Foundation to complete a one year study of the impact of adding brief psychological input into the local diabetes service (something I have previously blogged about). We reached the end of the year and applied for further funds to spread the impact, but were told that they do not fund follow-up data collection, or time to write up results for publication as this should be done in the course of your employment in the NHS. Of course we don’t work in the NHS, and have no funded time as part of a bigger contract or our contract of employment. So the Assistant Psychologist who was doing our data collection and analysis was only funded for the duration of the project, and has since moved on. The £2500 they did offer towards a editing our service user feedback into a short video, revamping our website, a press release and a small local event, requires a long report to release which is now overdue. I’m still trying to work out how to complete and submit the academic papers, but this takes time and means it has to compete against all the other demands of running a business.

Herein lies the big double bind: you need funded time to write bids and to write up papers to give the credibility to future bids, but unless you have academic tenure this time is not funded. For a clinician it competes with more immediate clinical demands, and for someone self-employed it competes with the tasks that actually pay the bills (in my case training, consulting, and up to 4 court assessments per month). There are also other things that demand my time and mental energy (committees, other forms of writing, http://www.clinpsy.org.uk, etc) and I sometimes even have a life outside work. Not everything can fit into the evenings and weekends, even if I allow work to expand into the rest of my life like gap-fill foam.

But I am not defeated. I am determined that the diabetes results will be submitted to peer reviewed journals by the end of the year. And I’m even more determined that what I have learnt about risk and resilience factors in adoptive matching will make its way into social work practice. I’ve considered all kinds of options including writing it up as a PhD, or a book, or a series of papers, or all of these things seeking funding as a grant, or a stipend, or a fellowship. I’ve tried to use my network to find potential funders. But nothing seems to come to fruition in a way that would allow me to have funded time to write it up with input from a statistician, whilst still having at least half my working week to run my company. Perhaps that is just asking the impossible, no matter how close to fruition the research is or how it will impact on people’s lives. I don’t know.

Meanwhile I keep asking the questions and gathering the data that will answer them, even if I can’t share the results. Innovate, evaluate, disseminate is simply part of what I do. I’ve got grand plans for what the next project will be, but I’ll save them for another blog.

BTW, if you have any ideas where I can secure funding, or want to collaborate with me I’d love to hear from you. Likewise if you can offer an overloaded clinician with big ideas a nominal NHS and/or university home, that would also be very welcome. It is twice as hard to bid for funds when you work in a small company that nobody has heard of!