Talking ChatGPT at DDD North 2024

/

A great audience here in black and white

Did my talk at DDD 2024 this morning. After a brief panic where it turned out that my HDMI adapter wouldn’t talk to the video system I managed to get everything working (thanks Warren for lending a working adapter). The audience was great and we had a splendid discussion afterwards. You can find the slide deck here.

First Man In at DDD

/

What Dall-E thinKs I look like. I’ll show the real picture of me at the session…

I’ve got the schedule for DDD North next week. I’m on first thing at 9:30 which will be fun. I’m talking about ChatGPT and my existential crisis.

Apparently they’ve had such a demand for tickets that they’ve had to add a bunch more. I’m really looking forward to it. Now I just have to decide which camera to take along….

Developer Developer Developer Videos now live

/

A few weeks ago I had a great time talking about making music with the Raspberry Pi PICO and Pure Data at the Developer Developer Developer Conference. All the videos are now online. You can find them on the conference YouTube channel here.

If you want to watch my video (and why would you not) you can click on the link above.

DDD North 2019 is coming to Hull

/

If you’re in the Humber area (or if you’re not and you don’t mind a bit of travel) there’s a splendid developer event happening here at the start of March. The Developer Developer Developer (DDD) North conference has a whole bunch of wonderful sessions and a fantastic venue (The University of Hull).

I’m giving a session too, all about Air Quality and Azure Functions. It’s going to be fun, I’m really looking forward to it, and you should too.

You can register here. It’s free.

Machine Learning at the Insider Dev Tour

/

If you were there, click through to the large version on Flickr and find yourself...

Today I had to get up early to catch a non-existent train. Of course, I didn't know it was non-existent as I got showered and hurried down my breakfast before dashing out of the house. I found out at the station. Apparently my train to Manchester was cancelled because it was "late leaving the depot". Which makes it sounded like they were trying to blame the engine itself. Anyhoo, after a brief panic I caught an earlier train that arrived later and made it to Manchester in time for my session. Which was great fun. 

I was talking about "Machine Learning", which is the tool that you use when you don't know how to solve the problem. As I said during the talk, there was a time when the thinking was that when we can get computers to play chess, we'll have cracked artificial intelligence. This turned out to be wrong. In fact, things that are hard for people (play chess, do sums very quickly etc etc) are comparatively easy for a computer but things that are easy for a human (get meaning out of conversation, walk down a busy street etc etc) are very hard for the machine.

However, this is changing. Computers can do difficult things now. And you can build these behaviours into your applications. I showed off some software that recognised emotion, and another system that could tell stick figures from fish, and all using C# and running in my lowly Surface Pro 3.

If you want to have a go at this, and find out more about Machine Learning (and you should) take a look at the labs pages for the tour which are here. 

Thanks so much to Microsoft for putting the tour together and Pete for making it work in Manchester. It was great fun and the audience were lovely. And, to make the day even better, my train back actually existed. 

Scott Guthrie at Azurecraft UK

/

Scott Guthrie is one of my heroes. Proper developer chops and the ability to deliver a really good presentation. I've seen him speak a couple of times and event had the pleasure to present at TechDays in the session before his. So I guess you could say I was once his warm-up man. 

Anyhoo, he's speaking in the UK on a couple of dates in early June. Once in London and once in Reading. I'm not going to be able to make it unfortunately, what with examinerish type things happening around then, but if you get the chance you really should go. 

Find out more and sign up here. 

MVP Powered Techdays Online next week

/

If you fancy some cracking free training next week you can sign up for TechDays Online. There are a whole bunch of splendid sessions from Microsoft folks like Mike Tautlty and Paul Foster, along with some really good MVP speakers too.  It runs over Wednesday 18th May and Thursday 19th. The whole agenda looks really interesting, but yo you can pick and choose what you watch.. You can sign up here.

If you do go along, tell them Rob sent you. And mention cheese. 

Careers Networking Event

/
Never underestimate the attractiveness of free food....

Never underestimate the attractiveness of free food....

We had our third Careers and Networking Event today. We had six companies presenting and it was standing room only in the lecture theatre during the presentations. I think we'll find a larger venue for next time. 

There were lots of interesting tales of local companies doing world beating things, but for me the most impressive thing was that most of the companies that came to see us had hired (and even brought along) ex-Hull graduates who now worked with them. 

This was a follow on from our Games event a few weeks ago. We'll be doing it all again in a year's time. If not sooner. 

Games Networking Event

/
Sandwiches and Networking

Sandwiches and Networking

We had our first ever Games Networking event today. Games studios from up and down the country came up to Hull to talk to our students and discuss the state of the business. It was interesting the way that a composite viewpoint emerged with each successive speaker. These are the main points that were made

  • Good qualifications are nice, but a great portfolio of things you have made is the thing that will really get you places. Particularly if it is packed with stuff you did "just to see what you could do" and shows off lots of different technologies.
  • Networking is key, and Twitter is one of the best ways to make contacts. But don't forget to go out and meet people too.
  • Game studios may come and go, but there is always a market for capable developers. And everybody that spoke was building their business, with the consequent need for new hirings in the future. 

You can find more details of the event and, if you are a Hull student sign up for the next one in February, here.

Oh, and Adam really, really, wants you to sign up for Global GameJam this weekend. 

Much Travelled Dev Day

/

If you've ever wondered if it is possible to travel to from Hull to Mons in Belgium, do a session at Dev Day 15 and then fly back on the same day I'm able to tell you that it can be done. Even if the day starts with a 90 minute delay before you take off and you arrive too late to make your ongoing rail connection.

Fortunately, because the people at Thalys are completely awesome, this was simply a matter of rebooking on a later train (at no extra cost) and then heading off. I got to the venue around 20 minutes before my session. You can't beat just in time delivery. I did my session, and then headed straight back for my train home, getting back just in time for bed.

Talk about a full day....

Thanks to Christophe for the picture. 

Thanks to Christophe for the picture. 

Anyhoo, the session was great. The audience was lovely, if a little bemused at the start, when I started handing out small cheese and big cheese prizes. But they soon got into the swing of things and much fun was had. Apparently it was captured on video, which will be interesting.....

Thanks to Olivier and the crew for organising the whole thing. It's actually a rather inspiring story. A bunch of developers wanted a conference in their part of the world and they just built one. They've got lots of delegates, sponsorship, a great venue in a cinema multiplex and splendid Microsofties like Andy Wigley coming along to give them sessions on the latest tech. Great stuff. A pleasure to be involved with. I'll be there next year if you want me. 

Although I might travel down the night before.....

Rob on the Road

/

Every now and then the department lets me go out on the road and do the odd talk. Perhaps I could have phrased that better.....

Anyhoo, I'll be treading the boards in a couple of conferences later this year. I'm not saying you have to go just because I'm presenting but......

On the 11th November I'm doing a session at Future Decoded in London. This will be all about my little programming framework called SNAPS. How I built it, what it does, and why it is such a wonderful companion to a Raspberry Pi running Windows 10.

SNAPS is a big part of my forthcoming new programming text (I just have to write the first three books first - ho ho) and I'm really looking forward to showing it off.

In December I'm heading over to Mons in Belgium to do a session at Dev Day 2015. Again the focus will be on SNAPS, but this time I'm going to take a peek under the hood and explain how you can use some SNAPS implementation magic to help make better Windows apps of your own. And there will be cheese. It's a promise.

Devices and Networking Summit Day 2

/

And so to day two of the summit which features more impressive tech followed by a nerve jangling train ride and late night return to Hull. Now read on....

Keeping Control of Security and Privacy in a World of Devices

In this session we had some talks on different aspects of security. First up was a discussion of techniques to address "man in the middle" attacks, where a bad person interposes themselves between you and the person you are talking to. This technique of intercepting and relaying messages is particularly dangerous in a world where people are happy to use the nearest open WiFi connection with no thought of security. Then we moved on to a fascinating discussion of how we can remove the faces of "innocent bystanders" from photos that we take when we are out and about. This is a surprisingly tricky thing to do and led to a great discourse about how privacy can be managed in the wild.

The Next Big Hurdle

The next session I went to took a high level view of the development process, with the goal of working out how to make properly useful devices, both now and in the future. This is something that we are not terribly good at today, but with increasing levels of complexity and a greater range of uses than ever before, it is important that we make sure that the devices we surround ourselves with aren't more frustrating for us than they are worth.

The talk was placed in the context of producing something really useful, a way of allowing blind people to navigate around cities on their own. The project uses bone-conduction headphones which relay navigation advice generated by a smartphone. The most impressive thing about this for me was the intent to build a system that works even when the user is doing something else. In other words the user would be able to carry on a conversation even while finding their way around.

Home Networking

I wasn't quite sure what to expect from this session, but what we got was some very good discussion of the problems faced when you connect households to the internet. Things have moved on from the early internet where a home had one computer and one wire to the router. Nowadays there will be many devices connected via WiFi and a huge problem diagnosing what has broken when things fail. I saw the results of an interesting study looking for reasons why home networks fail (a hint: it's usually the WiFi connection) and some very interesting ideas about moving the home router into the cloud.

Lunchtime Demonstrations

During the lunch break they had demonstrations of some really neat stuff. I wandered around with my camera and took some pictures.

This is a way of making 3D rapid prototype printing even more rapid. Rather than printing out a filled in model the printer produces a 3D outline. There are a number of technical challenges to address here. The printer needs a larger hole in the nozzle to print thicker strands, the output has to be carefully cooled as the models are built and the slicing software has to be re-written to drive the printer in a completely different way. Very clever.

These are Picco devices. Tiny miniature screens that you can use to send playful, or useful messages to folks you know.

Inside each tiny device is an .NET Microframework powered processor, a WiFi interface, SD card and display unit. The devices and their 3D printed cases were created to explore product development as much as invent new kinds of interaction. Very clever and great fun to play with.

 

This picture shows some prototype "Disruptive Display" devices. These provide a completely new kind of display medium that "remembers" what light you shine onto it. By using light sensitive transistors in the construction of display drivers you can make a display which can be triggered to retain the level of light that is falling on it at a particular moment. The way I saw this, it meant you could make a wall display that you could "program" using a video projector whenever you felt like changing the scenery. The display doesn't really "remember" anything, each individual pixel retains their setting. You could also use this to "paint" with a torch, and then flick a switch and create a new picture.

There were also folks showing off circuits you can draw, some amazing touch devices on flexible materials, devices that can create radio signals entirely in software, posters you can vote on and a networking system for servers that can reconfigure its topology dynamically  to match the use case. You can find details of all the projects here.

Micro Datacentres and Cloudlets for Mobile Computing

The final presentation of the day was from Victor Bahl, who talked about the way that the cloud is becoming part of the way our machines work. Heavy duty tasks such as face recognition reap a huge benefit from cloud computing. The results are obtained more quickly and place less of a load on the batteries in the mobile device. But of course to use the cloud your device must send the data to be analysed and then get the result back, which depends on fast data transfer.

In many applications it is particularly important to get responses back quickly. Introducing a few milliseconds delay into the loop can make the difference between life and death for a video game player.

Victor talked about the way performance can be improved by streamlining the protocols used to send requests and also how large monolithic data servers were not the complete answer to the problem. Instead he described how smaller numbers of servers, "cloudlets" are now being deployed. These provide high performance local resources and reduce network traffic.

I wasn't able to stay to the end of this session because we had to grab a cab to the station. I had a tight schedule for the journey back. When I organised the tickets I thought that an interval of twenty five minutes would be plenty of time to get from the Eurotunnel train to the Hull one. Of course I was making the assumption that everything would run on time, which turned out to be mostly right. But it did mean for an exciting sprint across a tiny part of London.

Thanks so much to Microsoft Research for Inviting me. It was a great event and it was lovely to catch up with some folks I've not seen in a while.

Devices and Networking Summit Day 1

/

Note: These are my recollections from my track through the proceedings and are a bit subjective - but of course completely right :). I hope you find them an interesting read.

The summit proper started today. The first two Plenary sessions set an amazing standard to get us started. Very thought provoking, not just in the context of what people are doing, but also why the are doing it. And what can go wrong.

Opening Remarks from Peter Lee

The first of the sessions was from Peter Lee of Microsoft Research. He described how research and development work together, telling the story of a research group who devised a technology that used an array of 9 microphones to focus on a single speaker in a noisy environment. When they saw this the marketing department first asked for cheaper microphones, then for only four of them, and finally an ability to work with minimal calibration.

The group duly delivered and their technology became part of the Kinect sensor, one of the fastest selling gadgets in history. And the punch line of the story was that the academic paper describing the techniques that were used was subsequently rejected by a journal as having limited practical application. Such is life.  

Research is hard work, and non-linear. Peter talked about another research area which is chancging the way we use computers, voice recognition. He described how for the first ten years of this century the rate of errors in recognition remained stubbornly at around 20%, despite the best efforts of researchers in the field. Then, suddenly in the last few years we've seen a massive improvement in performance down to an error rate of around 7%, to the point where conversational translation is now going mainstream.

And the outcomes of research are not what you expect. One popular use for speech recognition/translation turns out not to be about talking a foreign language, but simply being able to see what people are saying if you are hard of hearing.  Very interesting stuff, and a great context for the ongoing discussions about ideas and the application of them.

Feedback Control and the Coming Revolution: Raffaello D’Andrea

Raffaello D’Andrea has done some really amazing things. From helping to design the next wave of robots to help Amazon to fulfil orders to making incredible juggling drones. He talked about how machines can be made to learn and adapt, and that it is probably not fair for us humans to make fun of clumsy robots. After all, it took several years for us to lean how to walk and move around with out breaking the furniture.

The Kiva robots and the flying machines that Raffaello worked on are both built with an ability to learn by practice and then over their lifetime continuously adjust to changing as they wear. This makes for properly useful technology, building on the original feedback mechanisms used to keep steam engines running at constant speed under different loading conditions. 

Feedback is a trick where you take the output of something and use it as an input. In the case of the steam engine the idea is that the faster the engine runs, the less steam it gets, making it stabilise at a particular speed. Slow the engine down by making it work harder and more steam is injected to bring the speed back up. But the problem with feedback is that it can be fiddly. The wrong amount of feedback can produce a wildly oscillating system or no movement at all. And combining two stable, feedback controlled, systems does not always end well. Although - strangely - the converse is also true. Two unstable systems can be combined to create a single stable one.

Raffaello said that he considered feedback to be a necessary evil, in that it made systems work, but should be used sparingly and with proper understanding of the potential for instability. He also voiced concern about the possibilities for problems when connecting large numbers of systems together. From a mathematical point of view it can be shown that this will lead to massive, unpredictable events every now and then. And considering that the Internet of Things is going to lead to a whole new connected world, this is something we need to think about.

Functional Materials and Process Enabled Device electronics

This was a fascinating and wide ranging session made up of presentations from a researchers in a number of different areas. There were talks about polymer batteries, printing electronic circuits on flexible materials, energy harvesting and even weaving electronics into material. Any session that contains the phrase "Infrared spectroscopy woven into socks" has got to be a good one. And it was.

Rapid Device Prototyping

After lunch it was time for a session on rapid device prototyping. The starting point was TouchDevelop, a development environment that is a great place to write and share code. You can create TouchDevelop programs on any mobile device and share them all around the world. And now TouchDevelop has an embedded presence too. Your TouchDevelop code can be downloaded into an Engduino device.

The Engduino is based on the Arduino platform, providing a whole bunch of sensors and coloured lights and software libraries to control them. There's emulation of the as part of the TouchDevelop framework, which makes it very easy to create programs, and they are working on simple network deployment of the compiled programs.

If you're looking for a quick and easy way to start coding, TouchDevelop is a great place to start. You can start writing code on your phone on the bus, and then complete the program on your PC later in the day. The code is all stored in the cloud and is dead easy to share. And the Engduino is a great little hardware platform, with plenty of connections including leds, accelerometer, magnetometer, Bluetooth and an SD card. 

Organic Seminconductor Science: Professor Sir Richard Friend, University of Cambridge

This was the final plenary keynote of the day. The focus of the talk was science, nature and power, with a great discussion of how solar cells work, how to make them better (and print them), and the fundamental science behind energy capture. There are some fundamental physical reasons which limit the amount of energy you can capture from sunlight, but this session gave a great description of how we can make solar cells which are better, cheaper and easier to make and deploy. Really fascinating stuff.

We rounded off the day with dinner on a boat trip down the Seine which was lovely. Great food, great company, great technology. What's not to love?

Hello from Paris

/
Pointing the fat lens at the tower

Pointing the fat lens at the tower

Well, this is nice. I'm in Paris for the Microsoft Devices and Networking Summit 2015. A chance to look at some very interesting embedded stuff. In Paris. What's not to love. The hotel is just down the road from the Eiffel tower. So of course I headed over there.

I took the stairs

Not a bad place to reach your steps goal....

Not a bad place to reach your steps goal....

I love the Eiffel tower. Fantastic place. And great for photographs too. If you want to see full size versions of the pictures you can click through to the images on Flickr. The conference starts tomorrow. Can't wait.

I love these telescopes

I love these telescopes

View from underneath

View from underneath

Not a bad view from the balcony

Not a bad view from the balcony

Very nice reception

Very nice reception

Just got back from a great reception on the 11th floor of the hotel. There were splendid views of the tower from the balcony, and huge glass windows that let give you a great view of the surrounding neighbourhood - but proved quite hard to walk through. Even though I tried. Talk about making an impression.....

I'd love to have stayed longer, but it turns out that I'll be updating module descriptions for the rest of the evening. In Paris. Such is life.