Bristech came back with a bang this November, hosting a long a list of speakers to start a discussion on every techie topic from software to DevOps and Cloud Native to big data and machine learning. And of course there was a range of talks focussed on the soft skills that optimise our working habits; attendees had the opportunity to gain some pioneering advice.
The 2019 conference welcomed 500 attendees to the Watershed – and everyone definitely left with some serious food for thought. It was a great day bursting with learning, but if you missed it don’t worry too much, you can catch a few highlights from the day here! In true Bristech style, you also have access to all of the fantastic talks on the Bristech YouTube channel.
Steven Pemberton – Moore’s switch
Steven Pemberton is a researcher at the CWI in Amsterdam. His repertoire is pretty impressive, having co-designed the programming language Python is based on and a vast number of web technologies, HTML, CSS, XHTML, and XForms, to name a few.
Also, his second job was in Alan Turing’s old department, working on software for computer number 5 in the series he had worked on. Almost enough said.
When asked to badly describe his job, he says he is a ‘smart alec who tells 90% of the audience their jobs will be gone in the next 10 years – but don’t worry because we’ll all be dead before then if we don’t sort out this climate change.’
His talk is titled ‘Moore Switch’. It’s about programmes and languages designed in the 50s. Working in computing at this time is almost unrecognisable in comparison to today. Back then, it was cheaper to make three people check a new program through manually typing it out than to let the computer discover the errors.
During the talk, Steven compares the world’s first installed computer to the Raspberry Pi. He asks the audience how long do you think it would take for the Raspberry Pi to duplicate the work of the earliest computer running 24 hours a day for 10 years? The answer is 5 minutes.
Moore’s law predicted this trend – and it is still continuing today.
We now have the exact opposite problem: compared to the cost of a programmer, a computer is almost free. Hence the Moore Switch.
To tackle this problem, Steven talks us through research he is conducting called declarative programming. By his prediction, everyone will be programming declaratively in the future, so you should really check out his talk for extensive examples on this new theory.
Emily Rigby – Making 10% Time Effective
Emily Rigby is a Software Engineer Manager at OVO Energy – the energy technology company who wants to bring clean, affordable energy to everyone. But Emily’s talk isn’t a pitch for OVO, instead, she told us more about an emerging idea called 10% time.
What is 10% time you ask? Well, it literally means getting 10% of your work time back to do what you would like with it. This doesn’t mean a day off though; the idea is all about self-progression.
The term was coined by Google and it is essentially specific time to spend improving yourself, your skills, a product or just an opportunity to develop a line of new thinking. For tech companies, it is about recognising the talent of their engineers and giving them a chance to galvanise on their expertise.
It all links into the concept of motivating a workforce too. As we all know, money isn’t the biggest and best way to motivate people at work, so you have to go further than this. Emily tells us about the three key ways: mastery, autonomy and purpose.
She says that 10% time does exactly this – you’re there to improve yourself, you pick what to do and whatever you learn will have value in your working life.
In Emily’s talk, she goes through with us how you can deploy this in your own workplace, why you definitely should do this and explains the long-term benefits, so give it a watch to learn more about this exciting workplace culture shift.
Adam Harwood – Spatial interaction in XR
Adam Harwood is the Capabilities Team Lead at Ultraleap – formerly Ultrahaptics and Leap Motion, but the two companies merged earlier this year. Adam’s talk gives us all we needed to get to grips with spatial interaction in XR; his aim is to teach and encourage those considering work in the sector to get going with it all.
We learnt about the opportunity of spatial interaction, which is a much more intuitive mode of interaction than the most popular types in tech today. If we could take the things we are taught about interaction from birth into the virtual world, then surely this would be better for everyone?
When Adam says XR he means extended reality, he makes it clear this encompasses anything with a headset that gave your hands freedom of movement. This is because the main interest is interacting with the virtual world in the same way as you do in the real world.
The key to this hand tracking. Adam talks through all the exciting developments in this sector but is also frank about the limitations of this; he offers some great tips on how to design your product in the best way with these features in mind.
He tells us, “Seeing your hands in VR for the very first time is special – people look at their hands as if they’ve never seen them before.”
It’s crucial for the sense of immersion – this point is emphasised. In the talk, Adam discusses the do’s and don’ts of the tech, as well as demonstrating how Haptics can take the sense of presence even further.
Amir Safavi – All your estimates are awful
Amir Safavi, Tech Lead at Google, gives a talk that explains why ‘all your estimates are awful.’
To begin the talk, he asks his audience to put their hands up if they’ve ever had to make an estimate for a software development project, and if one or more of these ended up being wrong – everyone put their hands up. When asked if this was because the work took less or more time, about 80-90% of people said the latter. So, Amir set off in the task to explain why this might have been.
There’s a discussion on the importance of getting estimates right. He tells us, “a decent estimate really sets up the right expectations.” It also elicits the right support and temporary measures that may need to take place whilst the software engineers work on this.
The real problem here is that estimates of this kind are much harder to make than the estimates we calculate in other walks of life. If this is the case, why do we even bother?
Amir talks us through why they often are awful, speaking of reasons such as poor pre-work life preparation, overconfidence or imposter syndrome. And of course, Amir includes his best advice on how to avoid some of these pitfalls through examples of lessons he has learnt throughout his journey.
Ant Kennedy – A journey into Cloud Native Machine Learning
Ant Kennedy is CTO at Gapsquare – the company utilising data and prescriptive analysis to close the gender pay gap. His talk is a case study about this startup’s journey of adopting a Cloud Native (CN) approach to delivering machine learning.
This in-depth talk is a fantastic insight into how Gapsquare’s software has evolved; there’s a real sense of the journey they have been on, how they’re using the CN approach now and where they want to take it in the future. This is demonstrated through three key stages of development: investigation, training and serving.
Ant talks us through the various software and systems in play – there’s a discussion on where these sometimes fell short and examples where it really saved the day. The take home message about CN is that it’s resilient, manageable and observable. He outlines the real application of getting CN machine learning into production and in front of users.
For a detailed breakdown of the process, you’ll have to take a look for yourself!
Questions at the end also raise very interesting points of data anonymisation, open source collaboration projects and how businesses take it when you have to break the news of discrimination to them.
