Defence digital masterclass @ Barclays Rise

What it was:

The first Defence digital masterclass, with various speakers on digital themes, hosted at Barclays Rise, Shoreditch, on 4 May 2018

What I learned:

Artificial Intelligence:

  • The AlphaZero AI algorithm learns from blank slate, does not just replay earlier human learning
  • Human beings already have symbiotic relationships with dogs, possibly we can see future relationship with AI?
  • Key valuable skill at the moment is being able to analyse convolutional networks
  • Future AI will not be evil but may do terrible things purely in trying to help
  • The UK 4G network is expanding to 99% coverage
  • AI is difficult to define… Douglas Hofstadter ” AI is whatever hasn’t been done yet”
  • Loose definition of AI: systems able to perform tasks normally requiring human intelligence
  • General adversarial networks, train the forger and the tester at the same time. Can be used to optimise at scale and pace.
  • Neuromorphic computing: chips that mimic brain function by only firing the ” neurons” when relevant, enabling high computational power with low electrical power
  • The lower power consumption may be a clue, from a macro (physics) perspective, that the chip is using more brain-like processes. Current traditional AI architectures tend to use many orders of magnitude more power than a human brain.
  • ‘Centaur chess’ where humans are supported by machines. Have beaten the best humans and best machines. Implies that human-machine partnership and process may be key to military success?

Different scales / types of digital transformation in businesses:

  • Business model transformation,
  • Customer experience innovation
  • Operating model transformation

Thirty five percent of P&G products come from outside the company via its innovation approach
Many innovative large companies are sustained or accelerated through government contracts and investment; it’s not the case that innovation only comes from private investment
Military rituals are similar to agile rituals?  Could be a useful alignment.
DEFRA are using machine learning to read incoming documents, identify the name and address etc, and direct the document within department

What I will aim to do differently as a result:

  • Develop our model for digital transformation in defence, and to think about the technologies that underpin it
  • Think about business and process transformation as part of the model
  • Think about the working environment.
  • Make better connections outside MOD… we need to move to an ‘outside in’ approach

Business Ethics and Artificial Intelligence 2018

What it was:

CIPR written briefing studied as part of Continuous Professional Development, 27 February 2018:
Source: https://www.ibe.org.uk/userassets/briefings/ibe_briefing_58_business_ethics_and_artificial_intelligence.pdf

What I learned:

AI is relevant to Business Ethics – for example, how do you ensure your organisation’s values are being applied if decisions are being made algorithmically?

Potential risks of AI:

  • Ethics risk
  • Workforce risk (loss of jobs /skills)
  • Technology risk (cyber-attacks)
  • Algorithmic risk (biased decisions)
  • Legal risk (privacy / GDPR)

AIs may be accurate but nonetheless reflect human biases.

Open-sourcing may be important for openness and trust in AI systems – this could be especially true in Govt where trust is critical and keep source closed is less important.

“Explainability” is key to AI trust and to working alongside an AI partner.

AI work and contracts should specify responsibilities carefully – AIs cannot be held responsible for their behaviour!

Some practical steps organisations can take:

  • Meta-decision-making to ensure AI systems act in line with organisational ethical values.
  • Make sure third party algorithms adhere to ethical standards.
  • Establish a multi-disciplinary Ethics Research Unit.
  • Introduce ‘ethics tests’ for AI machines, where they are presented with an ethical dilemma.
  • Ensure staff have access to relevant training courses and communications re: ethical use of AI

What I will aim to do differently as a result:

In future I will:

  • Think about ethical and legal and other risks of AI projects at the design stage
  • Continue my learning and investigation of AI as applied to my organisation’s business
  • Consider whether ethics, compliance and legal teams should be engaged in AI projects.

Civil Service Alumni: Introduction to Artificial Intelligence

What it was:

Talks by subject matter experts on AI hosted by the Civil Service Alumni, at the Royal College of Defence Studies, London on 18 December 2017

What I learned:

AI is permeating business:

  • No single agreed definition of “AI”
  • ‘Narrow’ AI does specific tasks, as opposed to ‘General’ AI which can solve new problems
  • AI is already in use in hundreds of Google products
  • AI is increasingly in the plans of the forward-looking global industrials
  • Huge landscape of firms involved in different sectors

Deep Neural Networks:

  • The workhorse of modern AI is the deep neural network, where each layer of neurons is connected to the prior layer
  • Supervised versus unsupervised learning
  • AI performance is better than humans in many narrow case, using supervised data and deep neural networks

Conditions for success:

  • Many AI applications can be characterised and understood as input X leading to output Y to solve problem Z.  For example, medical scan data leading to disease symptom recognition to solve diagnosis.
  • A precondition for the successful deployment of AI is a clear definition of the system or process, and an understanding of what AI can do
  • A social technical approach, including human impact, is typically desirable
  • AI typically requires widespread digitisation of the system
  • Data requirements: Common definitions, high quality, accessible, sufficient size
  • Domain expertise is key

Impact on Government?

  • A viable, shared business model for both public and private sector is needed
  • Key challenges for government in terms of  capability, ethics, regulation, getting on front foot in terms of impact, and in nurturing the wider AI economy
  • There is a significant discussion happening now around ethics, bias, regulation

What I will aim to do differently as a result:

  • Think about ethics and regulation of AI
  • Think about impact of AI on Government, including potential oportunities
  • Continue my learning theme on AI!