This blog contains 100% Human-generated insight
That’s right. Before we start out this year, I want to be very clear on this point – absolutely no AI was used in the creation of this blog. This is 100% human. While 2023 was the year of generative AI, I want to start 2024 by denouncing it. Frankly, I don’t think you want to listen to a machine anyway. A generative AI is not going to tell you anything that you don’t already know*. In contrast, I hope this blog is going to tell you some things you don’t know. If I get it right, this blog is going to challenge you. And hopefully, this blog will give you the sort of insight that only a human brain can. So let’s be clear early on – generative AI has been nowhere near this. And maybe this is the first insight to take into 2024. 2024 is going to be the year of the human writer.
In 2024, statements distancing content from the use of AI are going to be increasingly important. It’s going to be more and more important to differentiate true human-generated content from the noise generated by machines. Why? Very simply, because words without insight are pointless. Words without insight are a waste of your time – you’ll learn nothing new. And generative AI in its current form simply cannot create insight.
Content generated by AI is predictable. It has to be – it’s predictable by its very design. If you choose your words based on what your training corpus predicts are the most likely words to see in response to a particular question, how can it not be anything but predictable? AI-generated words are by their very nature a statistical prediction based on vast amounts of input data. AI will give you great textbox answers – AI has literally devoured all the textbooks – but it won’t tell you anything else. Nothing learned through experience. No unique thought. Nothing new. Whilst this is not intended to be an investment blog, it is my humble prediction that in 2024 the current AI investment hype is going to come crashing to a halt [you may have noticed the lawsuits have already started with the New York Times lawsuit claiming copyright infringement and Gary Marcus and others pointing out how easy it is to get caught out if you think your AI is generating unique content (see here).
This blog’s not about AI – but there are some things you need to understand
But this blog is not about AI. This blog is about you. It’s about your actions in 2024. It’s about politics. And ultimately, it’s about what climate change means for our survival. But before we go on into all of that, let’s talk a little bit more about AI. Because whilst I’m denouncing it, generative AI was the biggest success story of 2023 and it’s useful to reflect on why that is and what it reveals about our current system.
Obviously, the technological breakthroughs of 2022 and 2023 represented a major leap forward in how AI could be applied to text generation. But the real revelation for me wasn’t the technology. The real revelation to me is simply how much demand there is for bland, predictable content. Generative AI has a value only because the content it creates is good enough to fulfil widespread business purposes – strategy documents, marketing blurb, policy documents, funding applications etc. Let’s recap – AI can’t create real insight, it’s not capable of doing that. Therefore, what is really surprising about all the hype is just how much content we collectively think that AI can replace. This indicates something really important – not about the high quality of AI, but about the poor quality of content! The real implication of this is that most of the content in the business universe lacks any genuine insight at all. The fact that we’ve become so excited about AI based on its ability to regurgitate sterile, bland, textbook answers should really be telling us something – that we’ve developed a system that so is overly bureaucratic that demand for text without insight is phenomenally high. You’ve heard the phrase “talk is cheap”. If words were cheap in 2022, then in 2023, all AI did was make them orders of magnitude cheaper.
Technology and Doxophobia
The second thing to consider about AI is how it’s trained and what this training approach might also say about society and debate in 2023 (in particular the current state of politics). AI training is based on a loss function. Simplistically, a computer program is tuned and manipulated to create text based on vast amounts of input data. If the output deviates from the training text it gets a bad score (high loss function). The parameters are then re-tuned to try and make the score less bad. This goes on and on until the output gets as close as possible to the desired result. At this point, the loss function is “minimised”. Then, from that point on the program outputs predictable results that are as close as possible to the input data.
Now, if that went over your head then that’s fine. We’re going to consider exactly the same thing but imagine for a moment that we replace the AI with a politician. Pick any politician – for our purposes, it doesn’t really matter what flavour of politician you choose. Now imagine that politician makes a statement. If that politician makes a statement that deviates from the “accepted” norm it might very well be plausible that that politician is instantly criticised by the press, leapt upon by social media and hounded until they either apologise, U-turn or are even cancelled on a more permanent basis. We can call this a bad or negative score or “high loss function”. Just like AI, it’s highly plausible that any politician or political party might feel they need to “re-tune their parameters” to do everything possible to minimise their loss function overall (you can see where this is going). If this were to happen and the loss function minimised, political statements would be much more predictable and much closer to the politician’s estimate of the “accepted” norm at the outset. I fear this is where we are headed – or worse, that we are already there.
Given the prevalence of social media, there is such little room for error today. Anyone in a public position has to have their wits about them at all times to avoid saying anything wrong. Not only this, smartphone video cameras are as ubiquitous as the telescreens in 1984, ready to capture for prosperity any slight misstep or slip of the tongue, while anything ever written or uttered on social media can be trawled through like a database of possible thoughtcrime. Against this backdrop, reasonable people are increasingly petrified of stating an opinion. This fear – the fear of stating an opinion – is called Doxophobia and in my view, this is a phobia on the rise. How often have you felt wary about sharing what you think in 2023?
Doxophobia and Debate
In my view, Doxophobia is having a real impact on debate. In innovation we’re taught the benefits of failure, the importance of trying different things. For example, according to the Harvard Business Review, innovative cultures are characterised by “a tolerance for failure and a willingness to experiment”. But it seems this mindset is lost when it comes to political discussion. There is little room to experiment with ideas, there is no room to test and learn and there is no tolerance for failure. Politicians (and increasingly normal people debating amongst friends) are expected to have fixed, unwavering convictions, to have tested everything before stating it and to have opinions that are as close as possible to the “accepted” norm – whatever that is at the time or within the particular group. To my mind, there are two key results of this trend:
- Firstly, reasonable people are increasingly worried about saying anything in a debate or contributing to debate publicly or on social media. Interestingly the extremists are still fine – they generally feel confident in their convictions and can find nice extremist echo chambers littered across the web. But without reasonable people to draw them back to reality, their views are becoming more and more extreme.
The Passage of Time
Writing a New Year’s blog has become a bit of a habit for me for a couple of reasons. First, the broad shut-down over Christmas gives my mind a chance to consolidate and scrutinise the year, and generally to wander to places it wouldn’t usually get the chance to go. Secondly, the New Year marks an important point of reflection for me. In the work we do, our own targets are measured on a calendar year basis, but more importantly, it’s also how our clients express their Net Zero target dates – headline years like “2030” or “2040”. The passing of each New Year is like watching the sand pass through an hourglass, the ticking of a clock or the relentless beat of a countdown. Time is running out. 2023 was officially the warmest year on record, at 1.43C above preindustrial levels. Some months of 2023 had temperatures exceeding 1.5 C above preindustrial levels, signalling that Earth is getting closer to the limits set out in the Paris Agreement (in 2015). The last six months of 2023 were described as “extraordinary” by Berkeley Earth with new “monthly records being set every month and by large margins”. And yet still global GHG emissions are rising (see here). Writing a New Year’s blog is my annual reminder that we have no time to waste.
Most Local Authorities in the UK declared climate emergencies in 2019. 2024 will mark the 5-year anniversary of those declarations. This will prompt an important review of progress. However, I have yet to meet anyone working in climate-related roles who has tried to argue that progress has yet been sufficient.
But the good news is that as well as marking the passage of time, the New Year also represents the hope and possibility that comes with a fresh start. It’s a time to reflect on why what we’ve been doing hasn’t been working to the extent we’d hoped, and a chance to think about what our actions should look like for the year ahead. But this brings us back to words. In 2021 Greta Thunberg famously said there had been thirty years of “blah blah blah” – “This is all we hear from our so-called leaders. Words that sound great but so far have not led to action.”
It’s true that words precede action. And when we hear words, we hope for action. But, at a time when AI can create words at virtually no cost, “blah blah blah” is actually on the rise. The only way we can differentiate ourselves is through action.
##2024: The year the country will decide Of course, 2024 is also going to be different because it’s an election year. The country will decide which party offers the most attractive vision for the country. A vision needs to be more than a collection of words. True vision needs to be backed by belief, conviction, commitment and action. That vision needs to be sold and repeatedly sold. And this is where I really fear what politics is becoming – lost, as it seems to be, in the noise of the “loss function”. Firstly, it’s not clear to me that anyone has fully articulated a positive vision yet. But more, I fear that where vision is stated, it’s once again just words. Just look at the toing and froing and distancing from ‘Green’ policies we saw as a result of the Uxbridge bi-election. That is a telling example of the loss function in action.
None of the problems our country faces can be addressed while at the same time keeping everyone happy. It is frankly impossible. But this is where true vision and the ability to articulate a vision is essential. It is often said that “people don’t like change.” But they do like change – people yearn for change, especially in the new year. The new job. The new relationship. The new house. The list goes on. The difference between those personal visions and the visions they’re currently being sold is that they can visualise that change, they can visualise how it will affect them positively, whilst also knowing it’s achievable.
Reaching Net Zero is all about making people’s lives better, but we need to be clearer about how it’s going to be better – we need to constantly sell that vision and we need to make sure that we remove the uncertainty of delivery so that people can actually believe that it’s achievable.
I don’t know what is going to happen politically this year – I have a suspicion it will be a Labour landslide, but I can’t promise it. I also can’t promise that if it is, Local Authorities will get the Net Zero funding they want and need. But the biggest risk we face in my view is simply the time spent waiting for the outcome. Our risk this year is being distracted by the electoral process and it’s hoping for certainty and funding that may not ever come. To me that only spells more delay and more inaction. ##What then should Local Authorities do? Local Authorities already have a vision for their region. Local Authorities already want action. But they feel their hands are tied by National Government. Yes, in some cases they are. But Local Authorities also can’t afford to wait for the stars to align at a national level before they just get on with it.
I believe there is a need, more than ever, for delivery mechanisms that are insulated from the uncertainty of the political cycle, that enable the acceleration of no regret options and bring in private finance to make delivery possible. That’s why for me, 2024 is all about Net Zero Partnerships – delivery partnerships between Local Government and the private sector that move beyond words and into action at scale. Net Zero Partnerships are about having a plan and delivering that plan. Net Zero Partnerships are about mobilising action from across the whole of the local area. Net Zero Partnerships are about maximising the funding secured from government but not relying on it. And most importantly, Net Zero Partnerships are about being able to look back in 2025 and say: “2024 was the year we made real progress”. Read more about Net Zero Partnerships here.
So, if you too are tired of the “blah blah blah” and want to secure genuine action this year, then let’s talk.
Laurence Oakes-Ash – Co-founder and CEO January 2024
Footnote: Just to be clear, my reservations concern the use of generative AI to create content, especially content masquerading as content written by a human. I remain a fan of AI and machine learning when applied to data analysis, for example, and we are developing a number of exciting projects in this area.