Exponential Growth with AI-Moments. Who needs the singularity?

Midjourney prompt: exponential growth with AI

We are in the Age of With, where everyone realises that AI touches our daily lives.

An AI-Moment is an interaction between a person and an automation, and these moments are now commonly boosting productivity or reducing our unwanted activities. Yet are we truly prepared to seize these opportunities as individuals, organisations, or society?

AI-Moments may be insignificant to us, for instance when a presentation slide is re-designed, or your car prompts a better commute route. These AI-Moments may be more significant when they determine every student’s academic grade [1] or rapidly evaluate a new vaccine [2]. AI-Moments are touching us all and they are the building blocks for imminent exponential growth in human and business performance.

Exponential growth needs AI-Moments that are ubiquitous, accelerated and connected.

Ubiquitous adoption of AI-Moments has already happened. It may be subtle, but everyone is already working with AI-Moments. Take this article that you are reading. An AI-Moment probably moved this up your notice list, created a list of people to share this with, helped your search tool find this article or prompted an individual to send this to you. As I am writing this piece, AI-Moments are suggesting better phrases, ways to increase effective impact, or improvements to my style [3].

Beyond the immediate pool of technology, AI-Moments are affecting how factories function through productivity tracking [4], changing call centres by replacing people with automated responses [5], or transforming our retail industry and high streets through online shopping. Take a moment to look at your daily routine or immediate environment to realise just how AI-Moments are already ubiquitous.

As you look around, consider how their adoption is accelerating in terms of quality and scale. This is because it is easier to create and adopt AI Moments. Applications are readily available that children can use to build AI-Moments that identify plants, recognise hand gestures, or detect emotions [6]. Monitoring satellite images for changes [7], recognising galaxies, or equipment analytics are all just as simple to build and adopt. Our most critical systems might require more robust solutions for the moment, but the acceleration of AI-Moment adoption is clear. AI-Moments that were not possible five years ago are now commonplace. They are, quite literally, child’s play [8].

Elsewhere, the first better than human translation with two languages occurred in 2018 after 20 years of research [9]. Applying that research to a further 9 languages only took 12 months [10]. This pace of change is universal. Google DeepMind solved a 50-year-old protein folding grand challenge in biology in November 2020, after four years of development and then mere weeks of training their AlphaFold solution. They are already now using that same model on diseases and viruses, predicting previously unknown COVID-19 protein structures [11].

AI-Moments are changing how we act, and their creation is changing how quickly we can re-act.

This creates a significant survival challenge, especially to organisations. An organisation that recognises, adopts, and accelerates AI-Moments across its functions has a distinct advantage over one struggling to do the same. Survival needs AI-Moments to break out of innovation or technology spaces, as rival organisations who deploy AI everywhere can act, re-act and improve faster while their competitors are still experimenting. Winners adopt and scale solutions better and faster using AI-Moments [12].

This will create the platform for exponential growth. First, we recognise AI-Moments are touching everything at greater pace and that they combine to multiply our performance. Then, as their pace expands, we realise that only AI-Moments can effectively manage this growth. People will find it too complex, or time consuming to understand, combine and exploit multiple AI-Moments. We will need AI to manage our AI with more AI-Moments.

AI-Moments are the common platform for exponential growth

Take the child’s app to recognise animals. The child shows the application a collection of cat photographs and the machine recognises cats. Show it a dog photo and it knows that it is not a cat, so we need another process to train dog recognition. The only way to improve recognition is through more cat or dog images, and even the internet has a limited quantity of cat photographs [13].

Instead, create an AI-Moment to recognise cats, then another AI-Moment to create synthetic cat photographs in new positions or environments. This is already a standard approach to train AI [14]. Using AI-Moments in this way exponentially accelerates learning as the only limit is the computing power available and not the quantity of cat photographs.

We can apply this approach to our current activities and processes, yet that creates a dilemma that will confront every single person and organisation: we will need more AI-Moments to manage, exploit and grow our performance. This will create exponential growth and, in turn, require more AI-Moments.

Our current concerns are around automating processes, replacing roles, or accelerating functions. They are “A to C” solutions, with success measured by how well an AI-Moment completes step “B”. Creating more complex flows is already normal, whether using another application to create them or copying someone else’s pattern to replace a familiar activity. These new complex flows effectively extend our solution from “A to n” with multiple steps in-between.

Automated AI-Moments will drive exponential growth, and will occur when existing automations are everywhere, accelerating performance and connections.

We are now on the cusp of significant transformation, where multiple AI-Moments interact, regularly in ways that we did not predict, expect or, sometimes, even request.

As an example, consider the routine of a typical salesperson. There are already solutions to automate office routines for meeting requests, room bookings and email responses. The first step is collating those automations into one “Get to Inbox Zero” AI-Moment, that involves a quick review of proposed responses and then responds: email replies based on your previous responses, all rooms booked, all requests sent ,automated prompts for more complex responses (that are expressed in simple language for the user to approve, “Yes, agree to request, use the agenda from the meeting last Tuesday”).

Then add in automated lunch reservations, travel tickets booked, hotels reserved, agendas created, minutes captured, presentations built, contracts drafted, and legal reviews completed. Include automated suggestions for new clients based on your current sales, existing targets, customer base, and market insights, with people identified to bring you together through an automated request that is already drafted in just the right way to get a positive response.

All these routines exist today in separate AI-Moments. Very soon these AI-Moments will connect and automate together.

There is often talk about the Singularity – the moment when machines will surpass human intelligence, and the idea that a single AI machine will achieve this superiority. The combination of AI-Moments does not need a super-intelligent AI, or General AI able to process any problem. It just requires a connected collection of ubiquitous AI-Moments, each replacing a small step of a larger routine. Each applies the rules of marginal gains and they come together to create exponential growth in potential. It may not be the singularity that futurologists predict, but its effect will be similar, as AI-Moments replace human activity in a way that surpasses human insight or comprehension.

This is the Age of With, and AI-Moments are the common units of change.

[1] A-levels and GCSEs: How did the exam algorithm work? - BBC News

[2] UK plans to use AI to process adverse reactions to Covid vaccines | Financial Times (ft.com)

[3] Introducing Microsoft Editor – Bring out your best writer wherever you write - Microsoft Tech Community

[4] This startup is using AI to give workers a “productivity score” | MIT Technology Review

[5] AWS announces AWS Contact Center Intelligence solutions | AWS News Blog (amazon.com)

[6] https://lobe.ai/

[7] As wildfire season approaches, AI could pinpoint risky regions using satellite imagery | TechCrunch

[8] Machine Learning for Kids

[9] Translating news from Chinese to English using AI, Microsoft researchers reach human parity milestone

[10] AI wave rolls through Microsoft’s language translation technologies

[11] https://www.deepmind.com/blog/article/alphafold-a-solution-to-a-50-year-old-grand-challenge-in-biology

[12] Building the AI-Powered Organization (hbr.org)

[13] https://en.wikipedia.org/wiki/Cats_and_the_Internet

[14] Adversarial training produces synthetic data for machine learning (amazon.science)

AI is the new Language

The technology of the Greek Phonetic Alphabet changed human creativity. Now, speaking the language of software and humans, AI will transform society and the software industry. In doing, AI will become our new language.

Language is essential to human culture. The fundamental difference between humans and animals is our ability to capture, communicate, and create using commonly understood language. Human evolution accelerated once humans could understand the sounds we make and capture those sounds in writing. Our ability to speak new ideas, write common laws, or read inspirational prose are the critical foundations of our modern global civilisation. 

AI is about to rock our language foundation. 

A comparison with the emergence of the Greek Phonetic Alphabet around 850 BCE is insightful to understanding how AI will now become our new language.

Around 850 BCE, the Phoenicians dominated naval Mediterranean trade, and their central location between Mesopotamia, Egypt and emerging Greek states enhanced their culture and commerce, with their influence extending from Afghanistan to Spain.

At this time, single rulers controlled all trade, contracts, laws, and decisions from their respective courts. Scribes formally recorded all deals, taxes, and judgements from the leader's court. Few people could read and write, and rulers seldom learnt this skill. To argue with a ruler, foolish in itself, was made more complex when few could understand the knowledge that established their rule. Nothing of worth occurred beyond the court walls as the court created all records. Therefore, widespread illiteracy led to the centralised power of the ruler, and scribes empowered this control.

A significant incentive existed to limit literacy. A scribe was well-paid, trusted and respected. Increasing literacy would diminish the worth of scribes' hard-earned literacy skills. In turn, rulers did not need their subjects to read their proclamations, only to obey them as subjects may disagree if they read the details.

Alphabetic complexity also made learning to read and write highly challenging. The Phoenician alphabet was consonantal, similar to all other alphabets at the time. To read, someone had to understand the discussion topic and remember many highly complex consonant clusters. While the consonantal alphabet was a significant leap forward from hieroglyphs and cuneiform, which used hundreds of symbols instead, it still took years of learning even to be basically competent at reading consonant clusters. 

Scribes needed to learn a word's written form and then comprehend the subject matter to translate that written form into actual meaning. The result was years of dedicated learning and apprenticeship before a scribe could capture in writing a trade negotiation, a new law, or an announcement from the ruler.

Between 850 and 750 BCE, the Greeks adopted the Phoenician alphabet and realised that Greek had fewer consonants than Phoenician, as today English has fewer consonants than Arabic. Reducing a complex process, someone then took the leftover letters to indicate vowels.

In Anaximander, Carlo Rovelli describes this moment,

"The many vocalic inflections of the same consonant - ba, be, bi, bo, and so forth - all rendered in Phoenician with the single letter B, could be distinguished as Ba, Be, Bi, Bo, etc.

It may seem a small idea, but it was a global revolution."

Indeed it was.

Greeks created the first phonetic alphabet, making reading and writing child's play. Learning the alphabet enabled someone to write the sounds they made in a way others could comprehend. They could deconstruct the sounds of others by understanding the same letters and joining them together. A sentence like, "A bird flies through the air" could be understood even if they had never seen the word "bird" before simply by saying each phonetic part. Once the reader said it, they would know that B.IR.D meant the word bird. 

The Greek phonetic alphabet was the first technology to enable almost anyone to record, share, edit, and understand the human voice.

The impact of this technology was immense. Anyone could hear the words of their rulers and, in response, share their ideas. Traders no longer needed a scribe to capture their negotiations and could escape the central court. Opinions could be shared, understood, and improved. Secrets could be passed outside the control of the ruler's court. Love letters could be shared. Propaganda could be published. This revolutionary technology empowered democracy, commerce, civilisation, medicine, literature, science, and, in turn, our modern information technologies.

The most significant change was that power was no longer in the hands of the few and could now spread through the reading and writings of the many.

How does this primary history lesson impact AI?

Since 850 BCE, the technology of the alphabet has underpinned nearly every significant revolution. Whilst other languages and alphabets existed, the need for mass literacy, or attempts to limit literacy, has empowered scientific discoveries, political revolution, or religious zealots. Even the language of mathematics, the other core language of modern society, prospered because the concept and ideas could be captured and recorded. We have democratised and made public our knowledge, our education systems have enfranchised everyone to understand our wisdom, and our civilisations have thrived.

The Age of Software Languages

Mass literacy empowered the masses and gave voice to their ideas until the 1970s, when a new language began to appear and evolve, the language of software. This new language ignored phonetics and, again, like cuneiform or hieroglyphic scribes, required specialist training and knowledge to comprehend. Someone without that knowledge could not deconstruct meaning using a handful of symbols; even if they could, it would often be language specific, reducing the detail one speaker could gain from another. 

The language of software rapidly evolved and changed, constantly changing, merging the languages that preceded it and parenting new languages. National critical infrastructure teams have gone through the ordeal of identifying old languages and the people who understood them to address security risks. A coder from 1979 may understand elements of some of today's software languages with their expressions and statements. Yet, even a simple quantum language like Q#, which includes quantum states and operations, would take much work to comprehend. It is, simply, another language.

The world has undoubtedly changed through the language of software. Its influence is felt direct through our interactions with digital devices, one step removed through managing the core services and utilities that power our cities and culture or indirectly through shaping today's societies. It is hard to dispute that we are in an age of software.

The language of software has also created new rulers and empires. MS-DOS founded Microsoft with an operating system that enabled PCs to work in a standard manner translating functions into system activities. Apple created a translator between human interactions and their computers using hardware mice and later touch screens. Google began with software to understand the knowledge of the internet and make it accessible.

The language of software is behind all of these empires, their scribes are well-paid and respected, and their language is not accessible to the masses, even if those masses have access to these empires through controlled portals. Guarding critical source code for crucial programs and, in turn, the valuable IP from that code is one of the highest priorities for any software developer seeking to monetise their code. The protection is not just to prevent security risks but also to guard the very source of their business.

Consider how Twitter/X recently reacted when Meta launched its social media posting solution. Its first response was to claim that Meta poached developers (scribes) and used the code (language) that made Twitter successful. Where that challenge will end or its legitimacy may be debatable, but software companies protect their scribes and languages just like ancient rulers.

AI threatens these technology empires (unless those empires control that AI).

AI is starting to break down these barriers of understanding and competition at an accelerating pace as we are on the verge of a new democratised alphabet.

Nearly everyone has become excited by Generative AI (GenAI), the suite of AI tools that generate and create artistic products like art, words, or music. Most readers of this piece will have experimented with prompting AI to create an image, asking an AI tool to write a summary of a complex document, or even generating a poem about a particular subject.

It is child's play to use these GenAI tools.

As well as capturing our imaginations by producing new content, recent GenAI tools have also broken down another critical barrier around ease of use and access. The interfaces that we use to engage GenAI has democratised access to AI. There is no need to understand data, storage, coding, models, or languages to access AI tools that generate immediate and tangible results. Their prompt interfaces and widespread distribution across devices, platforms and software have placed AI in the hands of the generalist rather than the specialist.

Creating and editing complex illustrations required access to someone with artistic talent and the ability to express the requirement. Now, "/Imagine Greek Philosopher holding a laptop inspiring a crowd" generates the image at the top of this post.

Previously, ease of access has been a significant barrier to adopting any new technology. The phonetic alphabet addressed this by enabling anyone with a canvas and a stylus to write. The results have been written on walls, hides, papers, rocks, metals, wood and screens ever since, but it still took centuries to revolutionise society truly. In 1820 only 12% of the world's population were literate. By 1960 it was 86%. Think of the possibilities if literacy at that level was achieved centuries earlier.

GenAI contains AI as both translator and creator to provide similar ease of access to complex tools and has equal potential to transform our technology usage at a pace measured in months rather than millennia.

An area of particular interest is AI-generated code. Here, GenAI writes code based on other code elements to create a software product. Creating this type of GenAI requires analysing large, existing volumes of code stored in repositories to learn, replicate and mimic software functions and services.

Right now, the outputs are relatively limited. It can generate code to extract data from a website, analyse the content, publish the results, or create simple webpages for human input to generate another related output. 

These programs are currently limited by the length of output generated by the GenAI product, by the number of code snippets drawn upon to create the code, and by the owners of the GenAI service to ensure that expensive compute resources are not consumed building complex programs.

Today's GenAI services still cost significantly in terms of time and cloud resources to deliver large outputs, yet that cost is rapidly falling. As a result, new opportunities will rapidly emerge as AI speaks software.

AI becomes the new language because it can speak the language of software and the language of humans through intuitive interfaces.

This blending of language already translates from idea to output, adds insight to our thoughts, and generates new solutions for our problems. AI can mutually translate native and software languages, improving both. Again, like the Greek phonetic alphabet, AI empowers people to understand and build more. 

It is removing the barriers that protect today's software scribes and empires. 

Rather than AI generating a document, spreadsheet, or presentation, we need to consider what happens when AI can develop the tools that make documents, spreadsheets, or presentations.

Imagine prompting GenAI "to build a word processor" or "create a program that lets me edit financial spreadsheets".

These are different from the questions currently answered by GenAI solutions due to the capacity reasons explained, but they are questions we can pose very soon.

Pace of Change - who shot JFK?

In 2017, I was privileged to witness a Microsoft team develop an AI solution that comprehended and analysed over 36,000 documents relating to the assassination of President John F Kennedy. Around 20 people took eight weeks to digitise the documentation, create an AI suite to understand it and develop another set of tools to analyse and visualise the understanding. It was an impressive feat made possible through intimate knowledge of available AI tools and the team's ability to exploit software languages.

Today, using AI, a single coder could generate a similar solution in eight hours. 

Continuing that pace of change, in another six years, that time could be reduced to 1 minute, even if the rate of evolution and adoption remains the same as in the last six years. We know this is false as progress accelerates exponentially rather than stays constant. We also know that humans agreeing on what they want will take far longer than the machines delivering the request.

However, based on current progress and pace, we will soon use AI to generate complex programs and create personal, unique solutions to generic tasks. We will use AI to custom-build a word processor that works as we wish, adds functions we need, and share it with others to enhance.

Part of me reads these words as pure lunacy. Why would anyone want to create a Word or GMail replacement when perfectly effective solutions and alternatives already exist? Will we continue to use our existing toolsets as we have for decades? 

The answer to why is a mix of personal customisation and a question of cost. On cost, The Software Alliance estimates that software piracy costs an estimated $45bn per year in lost revenue to software developers. This figure indicates that many people want to use software products but are unwilling or unable to pay for them. Many pirated software users would use an AI-generated alternative that was freely available with similar functions. Crucially, many currently paying fees for a software licence or application would be tempted by a freely generated option. 

Few people are loyal to a particular application because of its brand or name. They are customers because it completes a required task with an experience they appreciate. Consequently, software developers include customisation and personalisation features to improve the user experience.

Again, GenAI coding will enable users to introduce levels of personalisation unique to that individual. Backdrops, colours, icons, layouts, and features can all become unique and specific or changed at a prompt's notice. People can remove functions, redefine forms, and crucially embed the same GenAI into their solutions to learn how the application is used and prompt suggestions to improve it. User input to develop new features is vital to good software development. GenAI coding will let users bypass the need to engage with developers and write directly to the application.

Software developers will have to introduce similar features into their products to compete. Still, they cannot compete if users build their own software to save money. Even with computing and development costs, developing a bespoke service with GenAI may be significantly less than the lease or purchase of the generic application.

Building your Own Software empowered through AI (BOSAI), with AI speaking the languages of humans and software, will transform how programs are distributed and used. Once achieved, BOSAI will revolutionise the software industry that has created our AI services.

Today, we still need to reach the point where BOSAI can provide the functions or capacity required to replace large commercial software packages. It is also clearly outside software developers' interests to release a tool with GenAI that replaces their unique skills. Like the ancient scribes, developers face a challenge between empowering everyone with a new literacy in languages against removing their own legitimacy. 

This may prompt slower enthusiasm by some for these toolsets, yet it is also an ideal opportunity for disruptive entrants to the software market. After all, the software industry has been disrupting itself from its first days.

This is how AI becomes the new language.

There was no master plan for creating the Greek Phonetic Alphabet. Whilst people could see the benefits and simplicity of the approach, they did not act with an intent to change the world. They acted to make their daily lives easier to share and their daily chores quicker to complete.

So too, with AI. Software developers have created a tool that can replicate their own language and created a way of accessing that tool that can be shared quickly and simply. What was previously guarded knowledge with high barriers to understanding and use has now become easy to exploit. The master plan was not to change society but to make their daily chores of writing code and creating applications easier to achieve.

Yet, like the alphabet, the implications are immense.

Software has created today's society. Without software, our society struggles to thrive and prosper. AI becomes as powerful as the first phonetic alphabet by simplifying the language to understand and develop software. AI disrupts industry sectors and businesses by providing new opportunities and presenting different approaches that software previously answered.

The ultimate result of AI is to replace the rulers and scribes that established our software society by removing the language barrier between Humans and Software. As the new language, AI can empower us all to speak Software.

With this power in our hands and being child's play to use, what will we do with it?

Ethical Principles for Artificial Intelligence

To realise the full benefits of AI, we’ll need to work together to find answers to these questions and create systems that people trust. Ultimately, for AI to be trustworthy, we believe that it must be “human-centred” – designed in a way that augments human ingenuity and capabilities – and that its development and deployment must be guided by ethical principles that are deeply rooted in timeless values.

At Microsoft, we believe that six principles should provide the foundation for the development and deployment of AI-powered solutions that will put humans at the centre:

• Fairness: When AI systems make decisions about medical treatment or employment, for example, they should make the same recommendations for everyone with similar symptoms or qualifications. To ensure fairness, we must understand how bias can affect AI systems. 

• Reliability: AI systems must be designed to operate within clear parameters and undergo rigorous testing to ensure that they respond safely to unanticipated situations and do not evolve in ways that are inconsistent with original expectations. People should play a critical role in making decisions about how and when AI systems are deployed.

• Privacy and security: Like other cloud technologies, AI systems must comply with privacy laws that regulate data collection, use and storage, and ensure that personal information is used in accordance with privacy standards and protected from theft. 

• Inclusiveness: AI solutions must address a broad range of human needs and experiences through inclusive design practices that anticipate potential barriers in products or environments that can unintentionally exclude people. 

• Transparency: As AI increasingly impacts people’s lives, we must provide contextual information about how AI systems operate so that people understand how decisions are made and can more easily identify potential bias, errors and unintended outcomes.

• Accountability: People who design and deploy AI systems must be accountable for how their systems operate. Accountability norms for AI should draw on the experience and practices of other areas, such as healthcare and privacy, and be observed both during system design and in an ongoing manner as systems operate in the world.