I’m just finishing reading "The Age of AI and our Human Future”, by Henry Kissinger, Eric Schmidt & Daniel Huttenlocher, which discusses AI and the expected impact the technology has for society, commerce, geo-politics, security & warfare amongst other things. “Not since the enlightenment and the Age of Reason have we changed how we approach security, economics, order and even knowledge itself. Now AI will revolutionise how Humans navigate the world. AI is coming online in searching, streaming, medicine, education and many other fields and, in so doing, changing how we experience reality”. In my mind there is no doubt of the massive impact AI is set to have, but by anyone's standard these are bold claims.
If I were still a software company CEO or Founder, I would be asking myself how this plays out in my particular market and over what time frame? It’s clear I’d need to invest in integrating AI into my products, if I hadn't already. AI represents a challenge and a huge opportunity for any software developer.
But, if you are a venture backed software startup grappling with reducing your cash burn rate, to give you the runway to operate through the looming recession, should you cut into R&D (the very thing that offers differentiation for the future), or are there projects like AI-enabling your solutions that you simply shouldn’t cut, if you are going to hit the market at the right moment. Timing is everything in this instance.
“... in 2023 Chapter 2 of the Cloud era will go mainstream, especially all things AI. So, while tech companies will obviously need to cut their cloth according to tougher trading conditions, they should not do this at the expense of developing their Chapter 2 offering. In short, 2023 is likely to be a challenging year but also one that will bring great opportunities for those that are well prepared.” - Ian Spence, Megabuyte
It does, however, feel like we have been talking about this for a long time. I did a keynote for my previous employer, Exact, at their annual customer conference, in Rotterdam, 4-5 years ago. I focussed on the allure of autonomous driving, the ability for AI to diagnose medical conditions more accurately than your doctor (never having a bad day, and always up to date with the latest research). Whether AI could ever replace human capability in creative industries and finally, via video, an ominous warning from Bill Gates, that if the robots replace roles previously occupied by humans, you would have to tax the robots for each unit of productivity they would deliver. And we humans would be forced to focus on roles that “robots” could never do, involving human caring, kindness and empathy - for instance caring for the elderly.
“Most people overestimate what they can do in one year and underestimate what they can do in ten years.” - Bill Gates
Until recently, it seemed not much had changed as a bystander, many of these potential use cases don't seem to have moved very far forwards in five years. But maybe things are beginning to radically change, and 2023 is the year when AI starts to deliver on its promise.
“General AI”, where an AI could perform all the multiplicity of tasks that a human can cover, and hence replace us all (!!) remains a very long way off. However, Artificial Intelligence services that focus on a particular narrowly defined task, trained on vast datasets, and learning from their own activity, to improve over time are suddenly making rapid progress.
Different forms of AI, “conversational AI”
All of a sudden everyones talking about Chat-GPT and Microsoft's $10bn investment in the war for AI supremacy. Chat-GPT uses “conversational AI” to provide uncannily human-like answers to almost any question, having been “trained” on most of the content of the internet. Microsoft talks about using AI to make Azure an AI supercomputer for the world.
Time to 1 million Users :
Netflix 3.5 years
AirBnB 2.5 years
Facebook 10 months
Spotify 5 months
Instagram 2.5 month
iPhone 74 days
Chat-GPT 5 days
Google is in catch up mode, with rumours of their own “conversational” AI, called Sparrow, developed by their team at Deepmind. They are allegedly delaying its release to ensure that answers from Sparrow, cite the sources of the information used, something Chat-GPT doesn’t do - which may be important if we are to put our trust in artificial intelligence to answer our hardest questions. Right now the technology is still early in its development and from time to time comes back with answers that are clearly wrong.
“You may have read about the excitement surrounding the release of this new technology - and if you haven't yet had a play, I urge you to do so. So why is it worth engaging with AI now? Technology innovations are often hard to get a handle on as an executive. GPT is different. It democratises what have been opaque technical processes.” - Martha Lane Fox - Sunday Times, 22nd Jan 2023
There are a plethora of other similar services emerging. For example, Jasper.ai, which focuses on helping marketeers generate copy, blogs and articles in a fraction of the time it would take if they did it alone.
These AI are all single task solutions, and so perhaps for now, you as a human, may not be replaced by robots/AI, but you could be replaced by someone who uses an AI assistant to improve the quality and speed of their work.
These particular solutions take a conversational approach to human interaction, hence they offer people, far beyond the technology industry, an accessible way to begin to understand the potential power of AI - if you like a glimpse into the future, through solutions that anyone can use on the web. And these technologies are good, they are very good indeed. But they are still not “fool proof” and may result in inaccurate and outright wrong answers to your questions. So, if you're operating this technology, the ultimate referee or arbiter of content and answers generated remains a human mind.
“When you get an AI to write your marketing copy, will it still stand out from the crowd, and be differentiating versus your competition if we’re all using the same services to rapidly create content? Or will it lose its uniqueness and difference? “
Many other software packages already use AI and machine learning techniques (yours may be one of them), but are sitting inside and embedded in application packages, which are therefore not externally “exposed” to those using them. They do, however, substantially increase the potential of those packages to deliver more intelligence and more value to their users, particularly when focussed on a narrow task (narrowness helps), for instance within vertical market software. As a small example, Exact (I remain a Board Member) uses AI-style techniques to do bank transaction matching in its bookkeeping solution.
“One of the main ways GPT is going to do this is through automation of customer service. Think about all the times you've called a company and had to go through a list of automated options before finally speaking to a human. With GPT, businesses will be able to train the model to understand and respond to frequently asked questions. And that's just the tip of the iceberg - GPT can also be trained to generate personalised responses based on a customer's history.” - Martha Lane Fox
Around a quarter of the Members of Boardwave, the European networking community for Software Founders & CEOs, are already using AI within their products and using the technology as a differentiator. For many the technology is "under the covers" to improve their solutions breadth or accuracy - so not necessarily exposed to the user. This may be a quiet revolution, as it seems there may be a creeping industrialisation of AI, stealth-like invasion of our software solutions and the businesses that use them. Take this graph, from McKinsey's report “The State of AI in 2022”, it shows the types of investments businesses are already making in AI technologies, and the current levels of industrialisation of this new technology
European businesses spend 6x what they spend on Application software on Infrastructure Software. The challenge is that infrastructure software has traditionally been dominated by the US, whereas we, in Europe, have an inherent advantage in Application software, often by Vertical market, that embraces the cultural and legislative differences of our countries.
Case Study Rory Ceflan Jones:“The recent deal which has seen the German vaccine pioneer BioNTech take over InstaDeep feels like a big moment - for healthtech, for AI and for Europe as a contender in the battle to dominate this technology. BioNTech, the company behind the Pfizer Covid vaccine, paid £562 million for InstaDeep because it hopes its machine learning technology for drug discovery will cut the time and cost involved in identifying promising new medicines. The German company, founded by a husband and wife team with Turkish roots, is ploughing most of its profits from the Covid vaccine into finding new treatments for cancer. It signed an agreement with the UK government recently to trial cancer vaccines which could offer 10,000 patients personalised treatment.” InstaDeep is best known for Alphafold which can predict the 3D structures of proteins known to humans, this has insane potential commercial applications.
No doubt we should all continue to invest in developing our Cloud B2B application software, but we urgently need to also support the next generation of startups and founders in Europe who are embracing AI at their core. We have as much opportunity and right to build the next generation of AI infrastructure, as any other global region.
What about Geo-Politics?
This discussion still seems trivial when compared to the big issues discussed in “The Age of AI” that describes the need for nations and states to invest in protecting themselves from cyber warfare that uses AI, or even on the battlefield.
The US already has a lead and we shouldn't let them (or any other region) gain the same advantage in the next generation, which is where AI plays a pivotal role. Europe needs to nurture it’s AI startups, to avoid dominance commercially by the US, China, or Asia, the way the US dominated Web1.0 and Web2.0 with the likes of Google et al.
One thing, that is apparent, is top leaders, working on this, know how powerful and dangerous AI potentially could be once it’s released into the wild.
Elon Musk once said “I have exposure to the most cutting edge AI, and I think people should be really concerned by it…AI is a fundamental risk to the existence of human civilization in a way that car accidents, aeroplane crashes, faulty drugs or bad food were not — they were harmful to a set of individuals within society, of course, but they were not harmful to society as a whole.”
We need to own our destiny in AI-cyberspace, in protecting our national and regional interests, avoiding state actors meddling with misinformation in our political discourse and during elections, or launching massive cyber-attacks during times of conflict and tension. We’ve already seen the anxiety induced by allowing Hauwawei to sit at the centre of our telco infrastructure or the Trump administration's concern about Tiktok gaining a market position in the US. That's nothing compared to the overwhelming dominance that one country or region could excerpt over another if they could achieve an unassailable lead in the development of ever greater Artificial Intelligence.
We need to act now, investing in our own AI startups and infrastructure, creating a diverse, equitable and open environment. We must all be encouraged to understand the potential of AI, create the right regulatory framework that encourages innovation but also respects personal privacy and data security. The European Union is taking steps in this direction with their proposed Ethical Guidelines, which might just give Europe the edge it needs to establish itself as a leader in artificial intelligence. With continued investment, support and collaboration from both private and public sectors we can help Europe become a leading hub for innovation within the field of AI. With strong foundations in place for application software development alongside an ethical approach to new technologies, Europe has great potential to be at the very forefront of AI development. We must not be afraid of the potential of artificial intelligence, but embrace it, by investing in our future and building a platform that will enable us to capitalise on its incredible potential. Let’s create an AI-ecosystem that is diverse, creative and ethical, together we can make Europe a world leader in AI.*
*This final paragraph was written by Jasper.ai, none of the words are the author's own.
29th January 2023
"The Age of AI and our Human Future”, by Henry Kissinger, Eric Schmidt & Daniel Huttenlocher
McKinsey:December 6th 2022 : The state of AI in 2022 - and a half decade in review
European Union’s Publication 08 April 2019 : Ethics guidelines for trustworthy AI
Sunday Times : January 22nd 2023 : Martha Lane Fox opinion article
Megabuyte Newsletter article : Ian Spence - January 2023
With use of jasper.ai - an ai for marketeers to write or summarise blogs and articles.