Archive
Subscribe on LinkedIn
AI leadership needs practical wisdom
David Knott David Knott

AI leadership needs practical wisdom

AI ethics must be more than a set of policies, checked off through a compliance process or reviewed in a governance meeting. Ethics must include reflection, opinions and never-ending arguments about how we should live, and AI ethics must include patterns of thought and behaviour which we deliberately, and deliberatively, seek and acquire.

Making sense of AI is difficult, especially if we take the task seriously: that is, not just attempting to understand the engineering and mathematics of AI, or even its application and economics, but also its impact on our minds, knowledge, roles, relationships and all the other things that make up our lives. I believe that a philosophical tradition over two thousand years old can help, as long as we approach it thoughtfully.

Read More
While you’re worrying about Hackerbot 3000, don’t forget about Human 1.0
David Knott David Knott

While you’re worrying about Hackerbot 3000, don’t forget about Human 1.0

Let’s use the term Hackerbot 3000 to refer to all of those AI models which supposedly present a new and unprecedented level of cyber threat. These threats have been framed in a few alarming ways: AI companies have declared that their models are too dangerous to be released; governments have imposed restrictions and regulations; and AI organisations have revealed that they have made successful cyber attacks on other organisations (although that’s not quite how they describe it).

I like the term Hackerbot 3000, partly because it is silly (and, despite the seriousness of these threats and their consequences, there is something silly about the rogue AI agent cyber attack narrative), but also because it captures the impression that we are confronted with faceless, implacable and dangerous machines which, just like the T-800 in the original Terminator film, can't be bargained with, can't be reasoned with, don’t feel pity, or remorse, or fear, and absolutely will not stop . . . ever!

Read More
Good news! It’s all their fault. Bad news! It’s your fault too.
David Knott David Knott

Good news! It’s all their fault. Bad news! It’s your fault too.

Several organisations have recently made statements about cyber attacks which they committed using AI systems which they failed to control. The language of these statements is interesting, not just because of its technical detail (or lack of it), but in how different parts are framed.

The parts which describe the causes and impacts of the incidents talk in the passive voice, and ascribe actions to non-human actors and external parties: ‘The incident occurred . . .’, ‘The models identified and chained vulnerabilities . . .’, ‘a model accessed the Internet . . .’, ‘our third party evaluation partner . . .’, ‘a misconfiguration by an independent company inadvertently allowed . . .’, ‘some of the agents being tested had engaged in sustained, potentially harmful activity . . .’, ‘an AI agent took autonomous, unsanctioned action . . .’.

When we read these phrases, we might wonder whether any human beings work at these organisations at all, and what they were doing when the incidents occurred.

Read More
Software is an illusion: AI doubly so
David Knott David Knott

Software is an illusion: AI doubly so

The term ‘digital native’ was invented in 2001 by Marc Prensky, in a paper which argued that teaching methods should be adapted to suit students who had grown up exposed to digital technology, particularly ubiquitous devices such as smart phones. Since then, like many such terms, the concept of ‘digital natives’ has been used, misused and contested. Perhaps its most simplistic characterisation is that in which young people are simultaneously capable of holding multiple conversations with their friends, streaming a film, making a video (and, these days, consulting with their favourite chatbot), while their parents or older colleagues are struggling to set the timer on the microwave.

I’m not qualified to judge whether the original use of the term ‘digital native’ was valid, or what the implications are for education (although I will observe that every generation seems to commission research which proposes that there is something weird and different about this set of young people, while describing differences that may be attributable to older people forgetting what it means to be young, and younger people lacking the experience of being older). However, I do think that there is a dangerous assumption in the term, that people who are highly familiar with a set of consumer technologies are somehow more ‘tech savvy’, and have a deeper understanding of the digital world.

Read More
Warning: philosophers at work
David Knott David Knott

Warning: philosophers at work

I’ve bumped into a couple of conversations about philosophy this week. The first was in the Economist, in an article which claimed that philosophy graduates may have a better chance of getting hired then computer science graduates, especially by AI companies. It’s not clear whether those philosophy graduates are expected to teach the models, coach the executives, or advise the developers, but it’s nice that they have jobs.

On the face of it, I should feel positive about this development. For the last ten years, since AI (in its pre-generative, pre-transformer incarnation) started to be used in the companies I was working for, especially in data-rich decision making use cases, such as fraud and financial crime detection and customer needs analysis, I have believed that AI ethics was an urgent problem.

Read More
Failing fast is fun until you fail at scale
David Knott David Knott

Failing fast is fun until you fail at scale

Finding bugs in your code is like being betrayed by your own brain.

Sometimes the betrayal is blunt and obvious. The system generates errors or crashes outright. When you look at the code, you realise that you have mis-spelt a variable name (again), or left out punctuation (again), or called the wrong function (again).

Sometimes the betrayal is subtle and insidious. The system appears to be working properly, until you realise that the data in the database is drifting away from reality. When you look at the code, it looks fine, until you wrap your mind around the logic and discover that it is decidedly wonky. And this is logic that you created.

Read More
Rethinking phase three: your AI adoption programme is stalled, but your organisation is adopting AI anyway
David Knott David Knott

Rethinking phase three: your AI adoption programme is stalled, but your organisation is adopting AI anyway

I used to tell a story about AI adoption which went something like this . . . 

AI adoption in traditional organisations has taken place in three phases.

In phase one, prior to the public release of ChatGPT, AI was a specialist pursuit for specialist people, solving specialist problems with specialist datasets. AI was used within banks to detect fraud and financial crime, or in retail to manage stock levels and distribution. But it was not knowingly used by most people most of the time.

After the release of ChatGPT, AI was available to anybody who could frame an English sentence and type it into a browser. It seemed to become a general purpose technology which could be applied to any business process which needed to handle language or make decisions.

Read More
Performative reasoning
David Knott David Knott

Performative reasoning

How many of your reasons are comforting illusions? How many of the decisions which your organisation makes are based on well-ordered reasoning, and how many are simply surrounded with the trappings of reasoning in order to make you feel better?

Organisations make a lot of decisions. What products to buy, what products to launch, who to hire, where to invest, which projects to support and which initiatives to cancel. These decisions are particularly apparent in the field of enterprise technology, where we make choices about how to design, build and operate systems, how to organise resources and how to adopt and integrate new capabilities. The need to tell machines precisely what to do seems to require precision in our own thinking.

Because these decisions seem important, we feel that we should take them seriously, and be seen to take them seriously. When we are making purchasing decisions, we construct elaborate scoring criteria, invite bids and conduct extensive evaluations. When we are planning investments, we build detailed business cases, evaluate ROI and risk factors, and construct portfolios of change. When we are running delivery programmes, we create dashboards, produce reports and run steering meetings, so that we can respond to circumstances and keep everything on track.

Read More
Embrace *all* the fundamentals, of software, of models - and of humans
David Knott David Knott

Embrace *all* the fundamentals, of software, of models - and of humans

I sometimes try to persuade business leaders that they should get to grips with the fundamentals of the technology they depend on. This often manifests as an exhortation to learn to code, not because I think they will be great coders (although they might be) or because we need more coders (although we do), but because engaging with the practical reality of building and running systems is the best antidote to the confusion, bewilderment and mystification that surrounds technology. And because they might make better choices about technology partnerships, investment, organisation, sourcing and strategy if they knew more about how it all worked.

For most of my career, this advice has been focused on traditional software: procedural code, written line by line by humans. However, as enterprises are attempting to figure out how to make effective use of AI, I think it is necessary for leaders to get to grips with the fundamentals of three types of work that goes on in their organisations: that carried out by software, that carried out by models, and that carried out by humans. I regularly see people confusing the characteristics of these three types of work, for example, claiming that an AI model is ‘like an intern’, or that there is no more need for a code base of software because AI will either make all the necessary decisions or generate all of the necessary code on demand.

Read More
Keep thinking; it’s worth the trouble
David Knott David Knott

Keep thinking; it’s worth the trouble

The term ‘cognitive offloading’ precedes the current generation of AI products. It was coined in a paper written in 2016 by Evan Risko and Sam Gilbert. In that paper, the term referred to the externalisation of reasoning and memory, including practices as advanced as using Internet search engines to find answers to questions, or as basic as tying knots in handkerchiefs to remind yourself that you have something to remember. It suggested that, although the practice has been around as long as humans have been able to manipulate their environment, offloading cognition could impair the ability to reason and remember, or be responsible for more subtle effects, such as undermining people’s confidence in their own thought and memory. The paper concluded that more research was required, particularly into metacognition: the practice of thinking about thinking.

Ten years later, the need for that research seems even more pressing. We have more tools on which to offload our cognition, and more people who are using them to do just that. Sometimes this offloading is explicit and deliberate, such as when somebody asks an LLM-based product to produce a business strategy or write an essay. But sometimes it is implicit and incidental, such as when somebody asks such a product to draft an email or summarise a document for them. They might think that they are merely offloading the work of typing or scanning mundane verbiage, but their choice means that there are thoughts that they will not think.

Read More
Living with uncertainty: be ready to be wrong
AI leadership David Knott AI leadership David Knott

Living with uncertainty: be ready to be wrong

Are you ready to be wrong?

Let’s imagine that you have followed the first seven steps in this article about becoming an enterprise AI leader (and maybe even attended the course to be launched later this year). You have learnt the fundamentals of computing, examined and adapted your leadership style, developed your leadership team, figured out your values and ways to stick to them, built a robust supply chain, re-imagined your enterprise and found a practical route to implementation.

And then everything changes. A leading AI company releases a whole new category of model which promises capabilities beyond anything you have seen before. Or the same company burns through its investors’ money, fails to raise new funds and the models you rely on vanish from the market. Or powerful, highly optimised new Open Source models are released which collapse the price of training and inference. Or companies come under pressure to recoup their capital investments and prices shoot up.

Read More
Managing the change: don’t turn your experts into novices
AI leadership David Knott AI leadership David Knott

Managing the change: don’t turn your experts into novices

Technology people often forget the human side of change management. It is hard enough to design and build systems, integrate them, migrate data, configure infrastructure, and get everything working and stable. Technology can be erratic and unpredictable: who has time for even more erratic and unpredictable humans? Perhaps if we tack a training course and a couple of videos on the end of the systems rollout, we can achieve a bulk update of the human config files.

The error of thinking this way was demonstrated to me when I was working on a large scale merger between two banks. The technology choice was easy: we picked the systems of one bank, and then started planning how to migrate data from one set of systems to the other. Our designs were filled with data extracts and transforms, temporary integration layers, reconciliation and testing suites. We thought that our hardest problem was fitting all the migration work into a constrained time slot.

Read More