Jun 15, 2023 | Change, Governance, Strategy
There is no way around the fact that AI is now with us, and evolving at logarithmic rates. The unanswered question is ‘so what?
There are two extreme schools of thought, and everything in between.
On one hand we have those who are extremely wary:
# It will replace jobs, creating an unemployed under-class
# It will take away peoples rights to privacy, choice, and freedom, creating risk from baddies
# The buggars will take over, we become the slaves of some dystopian thinking ‘terminator’ machines.
On the other hand, there are those who see:
# Huge commercial and community benefits from the automation and efficiency AI brings
# Every platform change in the last 200 years from coal to electricity, horses to cars, vacuum tubes to integrated circuits, PC networks to the cloud, all delivering huge benefit. Why not again?
# The risks are manageable, and less than the benefits that will flow, besides, it is now an unstoppable force, so choices are limited.
Let’s first have some context.
We have been idolising AI from our earliest times, seeking assistance, advice and guidance from all manner of sources. The beguilingly named Ada Lovelace, daughter of Lord Byron wrote what is seen as the first ‘software’ for the Babbage machine in around 1840, with Babbage taking the credit. In 1943 the first paper that associated the neural networks in our brains to electrical circuits was published. In 1950, 73 years ago, Alan Turing wrote a paper called ‘Computing machinery and Intelligence’ which posed the ‘Turing test’. This remains the central question of AI: ‘When can machines think?
The term AI emerged from a 1956 workshop held at Dartmouth College, seen as the birth of modern AI. It kicked off research work in many corners of the scientific world. Google, Microsoft, Amazon, Apple, scientists, and many startups such as Deep Mind, now part of Google, and OpenAI the designer of Chat and Dall-E, significantly funded by Microsoft, have been working on this since the 90’s. The ‘T’ in ChatGPT stands for ‘Transform’ a patented technology breakthrough by Google.
This long scientific road led to an inflection point last November when OpenAI let Chat GPT out into the wild to see what would happen, and take the strategic ‘first mover’ advantage.
What AI is: the application of maths and software code that ‘teach’ computers to synthesise information and generate output. It is controlled by people, although even the scientists are not always sure of what goes on inside the black box of software.
What AI is not: Killer software and robots that spring to life and take over by killing and/or subjugating people.
How does it work? Statistics and probability, combined with huge computing power.
The probability of a ‘u’ following a ‘q’ in English is very high, the probability of that q being followed by any other letter is very low. The probability of that ‘u’ being followed by an ‘e’ is higher than it being followed by a ‘z’. And so it goes, letter by letter, word by word, progressively taking on the context in which those letters, words, sets of words, and sentences are reflected, such that the difference between a ‘party’ in the sense of a happy event, versus a ‘party’ in the political sense is clear.
Having sorted all that out, what are the things we should be thinking about?
- AI as an augmenter. A tool that can assist us to outcomes that are smarter, quicker, and more comprehensive than we might have reached on our own. The role of humans will not be eliminated, but it will be changed.
- AI as a broker. AI stands between us, and an outcome we may not know how to reach, but can be facilitated by AI. You want to write some code, now you do not have to be a coding whizz, AI can do it for you quickly, and with reasonable levels of success.
- AI as a magnifier. Every kid can have an IA tutor, every doctor an AI coach, every scientist an AI collaborator, this will lead to potential productivity growth, scientific breakthroughs, creative boundaries being busted, reduce death in wars. The downside is also magnified, there is always a flip side to be managed.
- Should we be concerned with ‘Synthetic Empathy’? we humans are social animals, what impact will this accelerating trend to isolation from physical contact and interaction have on our collective psyche?
- Blue Vs White collar displacement. Every platform change in our economies over the last 250 years have displaced blue collar workers, in favour of white collar so called ‘knowledge workers’. This one is different, it is the white collar knowledge workers, those who shuffle stuff around who are in the gun. There is no AI/robotics that can replace Albert the plumber, or Steve the sparkie. AI will change the support mechanisms they use, but will not change the simple act of fixing the leak in your bathroom or installing that extra powerpoint in the kitchen..
- Regulation. How can, and indeed should, we regulate, somehow. It is remarkably difficult to regulate something that does not exist. We have failed to regulate social media, despite with the benefit of hindsight, recognising the damage it can do. Compared to AI, regulating Social media would be easy, and we have failed to get that done. The problem is how do we go about crafting regulations that do anything at all beyond catching silly stuff, when it is in the outliers, and things we do not see other than with hindsight, that the real danger hides.
To answer the question posed in the header, it is my view that AI is an enormous avalanche of technical, cultural and digital change. We need to either get with the program, or get out of the way. If it is the latter, you will be consigning yourself to irrelevance.
This is not to imply it is all good.
AI does not have goals, it is not alive, it is just your toaster on steroids, so you can control it. AI is a tool, like any other, which can be used for good and bad, but indifference will lead to whacking your thumb with the hammer. The other thing about tools is that over time, they build equality and productivity.
However, the potential downsides are huge, the opportunity for evil have never been greater, but as the avalanche will not be stopped, you have to be in front of it to see and prepare for the pitfalls before you trip over them and are consumed.
Suck it up and enjoy the benefits!
Header cartoon credit: XKCD comic from the scary mind of Randall Munroe
May 30, 2023 | Change, Innovation, Marketing
In this new world of marketing, being reshaped by Artificial Intelligence, how should those concerned with the longevity and salience of their brands respond?
Innovate.
AI is really good at looking at what has happened in the past, but has yet to develop a crystal ball to tell the future. Marketers key responsibility is to tell the future, then shape the resource allocation decisions their enterprises make to best leverage what they think will happen. No future comes in a linear fashion, but AI can only reflect in a linear way, in response to the algorithms on which it was trained.
Strategise.
Strategy is a game of choice, where what you will not do is at least as important and often more so than what you will do. Again, these choices are based on what you think might happen, and as noted, these are never linear choices. Strategy in a world being homogenised by access to data will be more fundamentally important than ever.
Manage Communication structures.
Yesterday’s world was dominated by silos. The simple fact is that customers do not care about your silos, only how you deliver value to them. Enterprises have evolved hierarchical silo structures as the most efficient way to allocate and manage resources. That remained true until the mid-nineties, and most enterprises still have not got the memo. Today, even any hint of silos and barriers to communication internally, and more importantly with customers, will lead to a rapid and fiery death at the hands of data and its scribe, AI.
Remove marketing complexity.
The last 20 years have seen a multiplication and fragmentation of communication channels to customers and consumers, along with the inevitable silent middlemen and rent seekers who just siphon off dollars with little or no value add. The complexity of the choices and channels has created a situation where the analysis of the value of marketing expenditure is little short of a children’s guessing game. This is despite and partly because of because of the plethora of options and tools. The only way to address this complexity is to cut the gordian know and simplify, simplify, and then simplify some more. In other words, marketing focus driven by strategy. Easy to say, hard to do.
Generate attention.
The main game of being relevant in a huge homogeneous crowd is to first generate attention. You do that by being different, and being different with a big dose of energy being injected into the differences that are relevant to customers and consumers because they solve real problems, delivering them real value.
If you do all that, while leveraging the capabilities of AI, and digital systems generally, it will be your competitors that struggle, while you are ahead of the game.
Header cartoon credit: Tom Gauld
May 24, 2023 | Change, Innovation, Strategy
Dr. Geoffrey Hinton, often labelled ‘the Godfather of AI’ left his ‘home’ at Google so he could ‘freely share his concern that AI could cause the world serious harm’.
The idea of AI is not new. Philosophers and mathematicians through the ages have been speculating and writing about things we would now count as part of the foundations of AI.
Jonathan Swift gave the Lilliputians ‘the Engine’, Thomas Bayes built his probability framework that is still used every day, and Nicola Tesla built a radio-controlled boat for the 1898 New York exhibition controlled by what he called, ‘a borrowed mind’, and the first paper that recognised the potential neural symmetry with our own brains was publsiesh in 1943.
Alan Turing proposed what became known as ‘The Turing Test’ in 1950. This generated a surge of activity, culminating in three academics hosting a workshop at Dartmouth College in 1956. This workshop is now seen as the ‘kick-off’ of AI research, much as the Solvay conference in 1927 was the catalyst to the nuclear research that led to ‘the bomb’ in 1945.
From that workshop, great minds have been busily stuffing the AI equivalent of Pandora’s Box until Pandora, in the form of Open Ai’s ChatGPT opened it in November 2022, and let the whirlwind rip.
As in the fable of Pandora, one of the most enduring of the ancient Greek allegories, once the box was opened, despite all efforts, there was no stuffing the evils back in the box. Luckily, Pandora had also been created with beauty, intelligence, and tellingly, curiosity, which led to her opening the box.
This is as it is with AI.
There is the evil we all see, centred on the rapid destruction of the status quo in all corners of our commercial, private, and public lives. The beauty may be harder to see in the short term, but will become obvious in the longer term, after all, the last thing to escape Pandora’s box was ‘hope’.
If nothing else, it will be exciting, and for some, the source of significant new leverage that can lead anywhere.
Header credit: Dall-E with the instruction: ‘Create a painting of Pandora opening the box allowing the evils to escape, in the style of the ancient Greeks’
May 22, 2023 | Change, Governance, Leadership
The most valuable resource in every business is the time, talent, and energy of employees. However, as these are hard to manage, and do not appear on any balance sheet, they are often grossly under-managed or completely ignored.
By contrast, Capital is readily available, cheap, and readily returnable, and as it is recorded and easily managed, it consumes our whole management focus.
Dollars are really the only easy measure of compliance and outcomes. However, they are a poor measure of much else that actually makes a business successful.
Time, strategic and tactical focus, and employee engagement are our most valuable resources, but are universally managed poorly.
There is lots of advice from the digital ‘cheap seats,’ all of it obvious with thought, but equally challenging to implement.
Wasted time in meetings, unnecessary email, excessive double handling, unnecessary process delays, broken processes, cultural norms, and many others. All add complexity, ambiguity, and time, while obscuring accountability.
Employees start covering their arses by copying everyone, shifting responsibility, hedging on project delivery times, duplicating unnecessarily, and focussing on the trivial while ignoring the risky.
You have all seen it, and from a distance shook your heads and asked yourself:
“Why not a bit of common sense?”
These transaction costs consume huge unrecorded amounts of time and energy, which translates into commercial obesity.
Think about all this as the ‘organizational load’ that is being put on people.
This load is similar to being overweight. As complexity increases, we allow practices to creep in that adds obesity to processes. It is a bit like letting your belt out an extra notch.
As of November 2022, we have the most potent anti-obesity drug ever conceived in our midst.
Artificial Intelligence, trained on large Language Models arrived with ChatGPT3 leading the charge. The performance improvement between Chat 3 and Chat 4 launched in March 2023 is astonishing. Forecasting what Chat 5, 6, and 7, just around the corner will be able to deliver makes my head hurt.
Are you thinking about your obesity problem now??
Your competitor surely is!
May 1, 2023 | Change, Governance, Strategy
Strategy is all about choice: what we will not do is at least as important as what we will do.
When you dig a level deeper into the generic ‘will I or won’t I’, you come to the question of how do you make what are almost always difficult choices between options in the absence of full information.
At a top level, that choice is driven by two factors:
Cost structure.
There is always a cost to delivering product. Understanding the cost drivers associated with your product and business model is essential. You will then be able to make informed choices about how best to minimise those costs without compromising the value you deliver to customers.
Value creation.
Selling a product of any type depends on a buyer seeing the value created by their purchase as being greater than the cost of the purchase. ‘Value’ is a very personal term. The value of an expensive watch is not in the ability of that watch to tell the time, it resides in a range of psychological drivers that drive individual behaviour and choice.
Until recently, most of the costs involved were of a physical nature, now they are increasingly behavioural. Similarly with value creation, in the past it was the utility you got out of a physical purchase, but physical utility has been usurped by digital and emotional utility.
Understanding both is critical to success.
Apr 29, 2023 | Change, Innovation
30 years ago tomorrow, April 30, 1993, the public internet was born with the announcement by the European Organisation for Nuclear Research that they would publicly release the HTTP protocols that would change the world. These protocols had been created by Tim (now Sir Tim) Berners-Lee for use by academic and defence facilities and had been very tightly held. On April 30, 1993 they were posted on what would become the world’s first website and were to be freely available to all.
10 years ago, I posted a happy 20th message.
Leading up to that momentous release of the HTTP protocols, providing the initial foundation for today’s internet, the US department of defence had created the ARPANET (Advanced Research projects Agency Network) in 1969. The first email message being sent by Ray Tomlinson who first used the @ symbol to separate the recipient’s name from the network address to himself in 1971. By 1983 there was general agreement on the standards for communication on the internet, the TCP/IP (Transmission Control Protocol/Internet protocol), and in 1985 the first domain, Symbolics.com was registered, which remains live today.
Once publicly released, the standardised protocols saw a mobilisation of innovative resources from around the world, resulting in rapid development of uses and tools.
Mosaic, the first popular web browser was launched in late 1993, and later was renamed Netscape navigator, and Yahoo launched in 1994. Microsoft launched their competitive search tool Explorer in 1995, later incorporating it free into Windows, leading to the move by the Clinton government to take action under the antitrust laws in 1998, resulting in an order to break up Microsoft. This order was later lost on appeal, significantly due to the evolving dominance of Google as the preferred search engine. Amazon launched in 1995, Google in 1998 and amongst the wave of tech IPO’s in 1999 was Napster, the first peer to peer file sharing service.
From the launch of Wikipedia in 2001, we again had a wave of launches, most of which failed, but a few became the unicorns that changed our lives, Facebook 2004, YouTube 2005, iPhone 2007, Instagram 2010, and so it continues.
The most recent inflection is obviously the explosion of AI tools since the release of ChatGPT in November 2022.
If you extrapolated from this birthday out to the next milestone, the 40th, the only thing we can say for sure is that you would be wildly, massively wrong. That happens every time such an inflection point is reached. Extrapolation is useless, instead we need to experiment and innovate, a continuous process that will take us in completely unpredictable directions.
I hope I am around to see it.
