BBC North East Technology Hub Meetup
BBC Newcastle was built in the 80s and was custom built for television and radio for BBC Radio Newcastle and TV bulletins and the BBC have spent a year planning and doing a refurbishment of the building to make it more ergonomic and invest more in the site, the result of this is a new space which takes place of the BBC club and bar.
There are many product areas within the BBC, they had traineeships years ago to offer learning on how to use the web to get the skills needed for online and enabled opportunities in this space. There is a lot of investment here in the north east to bring the product and technology group here, they have huge plans with a seventy strong team by the end of the year and bring in the talent to work on their products and services.
BBC North East Technology Hub Meetup is an opportunity for the BBC to share their space with the tech community, they have people working here with an engineering background and experience of tech hubs and building them. This space will be more data related for data platforms and product and getting the BBC part of the wider community. They have a flexible and adaptable way of working, they are doing data mesh and looking at raw data ingest and data specific and the hub is aiming to grow beyond the seventy people they are planning.
BBC Research and Development also is represented, they had decided to invest in Newcastle again and the BBC needed someone senior to make it happen. They had no budget and no team to make it happen, just need to make sure everyone moves in the same direction. Knowing where things could go if you know and what would work for technology people rather than journalists, so need to have dual monitors or enough WiFi points. They got to create a north eastern technology hub that is based in Newcastle, there is £25 million going into production and then have built a building and a space where people can be productive and make BBC the first global public service digital organisation, built around the quality of their programming, product designs, implementation and values and be great corporate citizens and be part of the community that is already here.
What does success look like in a year and a half? There are ten product areas and have started with data and want to know what areas will work, this needs to represent the areas they have, want people in the team to experience other product areas with opportunities to grow with a range of products and services. BBC has a vision and mission in the digital space and want to make Newcastle part of this. They have a distributed model with people around the country but will all be working on the same mission to bring the BBC to you and how to bring a more personalised experience to you including the podcasts you listen to, you get the best of what you listen to and BBC Newcastle will be a big part of that. Grow the products and the data mission, they want to own their own data and not have a third-party, this is one of the products but there will be many more coming.
Are there plans to build a platform or product? iPlayer, Sounds to News and Sports and do partner with other broadcasters, they have a suite of BBC services and want to bring this together, this content it surfaced to you regardless of what platform you are on, what you are listening to or what are you reading. How they can build upon their architecture as at the moment it is fragmented, it is a program approach but not "that" program approach.
Will this space be available to other groups? It will be as long as it can be done in a secure way, and they are part of the Newcastle and Gateshead initiative and want to be part of the reciprocity of this.
How do they approach retention of staff? What are they thinking about? What's important is that people feel it is a nice space, they feel they can have a career here and things they can work on and can grow and develop. Be part of the BBC vision but allow people to work from home and maintain their work-life balance and are flexible to meet the needs of their employees where they can develop and grow. The culture is good, people's personal life is a factor where it is adaptable with a hybrid approach and is centred around culture. Engineers have time to themselves to learn, can go to conferences, BBC is very people driven and make it a great place to be and is a very different place to work and are able to attract people as are a value driven organisation and make it truly welcoming, how their staff is treated or how this is reflected in their output. Work that is socially useful and that you will be thanked for, they have the opportunity to sit in the radio studios and in the gallery and get to experience what it is like to be in a broadcasting situation, not just a space where people are separate, and people can sit wherever and work wherever you want and can collaborate. They have had hackathons.
How does the BBC develop their platforms? They have a product and area they are focusing on and are going to be using AWS but it depends on the product and area, is based on what doe the team want such as agile practices, BBC allow for economy for tooling to use whatever methodology they want, the demands are flexible and adaptable depending on the needs of the product. They have inhouse skills where needed, they are organised in functions with product, UX, engineering, product and data and come together and are led by their functions and projects have a multidisciplinary team at all levels with the product group structured this way and may have different interfaces with different teams.
Are there elements of open source? BBC do use open-source tools when and where needed but if they need managed services, they will do this and will also work alongside other consultancies for products such as working with raw data on AWS but will contribute to the projects they use.
How do they feel about competing with other products and how does this influence development? They have certain values they are uniquely placed to give and do have partnerships with other platforms, how can they improve the user experience and easy of use, bring together the content as they are one of the few that brings together video, audio and text. BBC have an established set of services so if they wanted to do something different this would have to be assessed to make sure it supports and doesn't interfere with other experiences, they can assess the impact, but they can't just do what they want but they can propose things and have this assessed. Processes and design patterns can be similar to other products, but they do what is best for them in terms of their mission.
Digital Future Daily said the breakout hit of the spring with what sounded like a real collaboration between Drake and the Weekend that uses AI generated versions of their voices and has been listened to thousands of times. What do we make of a world where this is commonplace and what will happen to the technologies that we label machine learning and artificial intelligence?
Terminology, treat AI and ML as just those two letters to be divorced from their full versions. What has happened where these have gone from not being used much to being available to many different people and are filtering into every part of our lives. We have a range of highly trained models that can do word and pixel probabilities to give us something that we ascribe meaning to. They are not sentient they are just highly trained models that can do something really useful. They are not being controlled or regulated and are being developed by companies with a range of agendas for future profit and without any proper control or societal influence on how they are deployed.
These tools are extractive of everything, they require raw materials to create the GPUs, need vast amounts of electricity and extract value it all sorts of different ways, but is presented as a magic genie in a bottle. This is the illusion we maintain when we engage with this technology, and it is such a good story we fall for it again and again. There are biases within the data, such as professors are the “he” in a sentence as it has had input into it biased data as the statistics tell it to. The models are producing this output due to the information that is input and they way they were developed and should not fail to recognise this. Concerns of the long-term philosophy that you have to think about the long-term benefits of innovations and if it is of enormous value to society. Projecting these technologies into the way we work today and the potential of the things we work with now is enormous. We have some amazing tools to support creativity, we are going to be using this in products we have now, it will just work and be incorporated in it, you'll get something you can start with that you can improve at least. We are good at compromising if it is quick but below par compared to something that takes longer and is better.
ChatGPT was used to design a board game to combine ticket to ride and another game, and define the game clearly which became a discovering object game and refined this to be a discovery of elements of the periodic table and at the end had a reasonable specification of a game that had be co-created with an AI, they were bouncing off the ideas without the other getting tired, it won't stop and won't get bored. GitHub Copilot and Chat GPT were used to scrape a website to get the information to dodge days of graft and made it feel like flying and this is using tools today, they are using the tool to get the job done and it does this for them.
Could use these tools to clear away some of the tasks in their lives and can be a well-resourced companion, no-one has lost their job to ML, someone has to take this job and given it to an AI. Where are these things going to be used? This could be the metaverse, how could you pull a lobster from your backpack - what is the UX for doing this to get the infinite object space of things you might describe. Do we want to spend half an hour doing this, this is going to be built using AI and ML so you can just ask for what you want and it will just be there and have ChatGPT style interfaces and want our metaverses to be different, you can be in the same virtual space but your rooms don't have to look the same, you should be able to change anything especially in things like neurodiversity, so will need some kind of automated help.
There is more to AI than foundation models, don't disregard the other areas as they solve another class a problem and might not necessarily go beyond this. There is a 10,000 word essay from Steven Wolfram about how ChatGPT works, sometimes it didn't work and the parameters needed to be changed until it works. Transfer the process of doing something without understanding what is going on, if AI and ML reach the limit of human understanding they cannot transcend the limitations of them, they won't become Artificial General Intelligence. The ability to compress knowledge and use it is when everything changed, main use-case for ChatGPT is to process the large amount of information, this is the phonetic alphabet for information storage.
User Centaured Design rather than User Centred Design. Can treat a computer as a collaborator and the way we interface with a computer changes, asking the right questions and acting smartly with the answers and there is a Partnership on AI.