Do flamingos dream of electric sheep?
From Ian Forrester ?? , our CEO.
And so the keenly awaited dawn of AI is upon us. Publications and platforms are awash with content expounding on this new phenomenon; AI is being heralded as the greatest driver of productivity since the arrival of the computing age but is also under fire from several quarters. Naysayers warn of the risk of AI taking human jobs, while doom mongers decry the singularity-induced creation of super-intelligent killer machines. But who is right? Will this technology result in delight or disaster??
No one, not even the most powerful AI systems known to man, can predict the future (yet!). But, as CEO of an AI-powered company and avid reader of the latest thinking on AI from around the world, and with AI seemingly the only thing spoken about at Cannes this year, it’s time I put forward DAIVID’s view on the subject. After all, few topics are as emotive as AI, and the understanding of emotions is at the core of DAIVID, so this article feels apposite.
WHAT IS AI?
First of all, for the uninitiated, confused or just plain bored, let’s discuss what AI is. And who better to ask for a definition of AI than the current poster-bot of the space, ChatGPT. Upon asking ChatGPT to explain artificial intelligence, it returned this response:
Artificial Intelligence (AI) refers to the simulation of human intelligence in machines that are programmed to think and learn like humans. It is a broad field of computer science that focuses on creating intelligent machines capable of performing tasks that typically require human intelligence.
AI systems are designed to analyse and interpret vast amounts of data, identify patterns, and make decisions or predictions based on that data. They can process information, learn from it, and adapt their behaviour over time, improving their performance and accuracy.
This is the crux of it: machines which simulate human intelligence; they learn and make decisions. No wonder AI gets such bad press; it’s not a huge leap to imagine how such technology might be used for nefarious purposes, and its ability to learn sends shivers down spines. “If it can learn, will it one day become more clever than humans? Will it become conscious? What if it decides that humans are no longer worth keeping around?!” And so on.
AND WHAT IS IT RIGHT NOW?
But hold on! We are a very long way from killer machines extinguishing humanity and before we consider that possibility we must first explore what AI is right now.
Firstly, we must realise that AI isn’t a singular thing, like Skynet or God or an individual program capable of doing everything. There exist today multiple AIs. These AIs are currently either Applied AIs or Generative AIs. Applied AIs ingest a specific set of data in order to answer a question or fulfil a use. The more data they ingest, the more broad their knowledge-base can seem (take a bow ChatGPT) but there is always a limit to that knowledge-base.?
Generative AI has similarly ingested a specific set of data, this time vast quantities of images and/or videos. It then generates new images or videos based on the user’s prompts, taking the ingested images or videos as its base. What it creates is ‘new’ in that it is recombinant. It combines things which have come before in new ways but does not create from scratch.
MADE IN OUR OWN IMAGE
Herein lies one of the greatest challenges of AI: it’s only as good as the data upon which it’s based. As so much of the world’s digital information is laden with bias and mistakenly or even deliberately misleading ‘facts’, a system based on this data is fraught with issues. As a result, concerns of the potential to use AI to spread misinformation abound. Even if there is no malicious intent, an article written by AI and published may well be full of factual errors.?
Yet when the data upon which the AI is working is intelligently composed and the AI is given strong direction, AI can produce magic. A perfect example is a recent MIT project in which AI discovered a new antibiotic that is now successfully killing antibiotic-resistant bacteria.
The AI succeeded by screening more than 100m chemical compounds in a matter of days and identifying patterns which were indistinguishable to humans. Having established the existence of these patterns, human researchers are now setting about explaining them, and in so doing are expanding the human race’s understanding of molecular cell biology. In this way, AI and humans are acting symbiotically, with the AI unlocking doors through which human understanding can step. It is this combination of AI computing power and human rational thinking and interpretation which provides the opportunity for an AI-powered society which is massively enhanced in many fields.
AI MAKING CONTENT
And what of generative AI? Its ability to create content at scale is being touted as a potential creative director-killer within the advertising industry in which DAIVID sits. Does AI have the power to totally supplant human creativity? By this author’s estimation, no.?
领英推荐
Creativity is the result of the merging of disparate experiences, experiences which are infused with human interpretation and emotions. These experiences come together to create something which previously did not exist. For AI to be able to create everything which humans can create, it would need to ingest every single human experience. Everywhere. All at once.?
While some point to smart speakers as the AI Trojan Horse, potentially giving AI a window into human’s everyday lives, this is still a very far cry from being able to see the world through 7bn people’s eyes every moment of the day. Not to mention the perspective given to each human by their unique upbringing, the upbringing of their parents and their parents’ parents and so on. The result is that humans will ALWAYS have a unique proposition to bring to the table.?
HUMANS AND AI WORKING TOGETHER
Humans need not worry about being replaced by AI; they should embrace this new technology as the facilitator of creativity that it is. Creators who use AI as another tool in their toolkit will find their creativity enriched. This is precisely DAIVID’s approach to the use of AI in creativity; we augment creators and set their creativity free.
And what of the evolution of AI? Quite apart from the relatively mundane concerns around job displacement, bias, discrimination, a lack of transparency and privacy, there are murmurings about AI eventually causing the end of human life on earth. This line of thinking starts with the creation of General AI. This is an AI which doesn’t have a specific focus but instead can turn its robotic hand to literally ANYTHING.?
The creation of such an AI would require a general dataset to fuel that learning. And herein lies the problem. Such an AI would either need to learn from existing knowledge or would need to conduct experiments to generate the data needed to learn, but there are significant barriers blocking the way of both routes.
NO GENERAL AI - YET
There simply doesn’t exist the body of knowledge that General AI needs to teach itself. As we have discussed, public sources of information are riddled with errors and inaccuracies. Looking at non-public knowledge, even if the AI had access to all academic and industry studies ever created, there is a lack of consensus in virtually every field of scientific discovery. Even when consensus is seemingly found, as in Newtonian physics, new knowledge and advancements come along and either sideswipe or augment the old establishment (enter quantum physics). Even if there were a universal, complete and 100% accurate training dataset, the AI would still need to understand the unsaid unknowns: things that humans take for granted due to common sense. For example, what if the training data talked about a roller skate race? The AI would likely think this is a race between roller skates instead of humans wearing roller skates. Common sense is the aggregation of decades of cultural assimilation passed from generation to generation. The mind boggles as to how this could be structured, categorised and codified to train an AI.?
Another option is that AI conducts experiments to generate the knowledge it needs to create General AI. Yet imagine the immense scale of experiments required for AI to learn everything about everything. The resources and sheer space required for this set of experiments is mind-boggling. And these resources and space would need to be gathered by an entity which currently has zero means of actually doing anything. All AI can do right now is think and there’s no way to think your way to conducting experiments. Even in a future world where the AI is connected to real world things (e.g, cars, drones etc.) it’s difficult to imagine a world in which it has the connections with sufficient breadth and dexterity to teach itself EVERYTHING.?
And so, the creation of the singularity, in which recursive self-learning rapidly leads AI to 1. Become super-intelligent. 2. Deem the existence of humans to be contrary to its goals, and 3. End all human life on earth, seems quite a long way from reality.
I agree that AI will transform humanity but in my estimation this transformation will manifest in a myriad of positive ways. Kant declared that there exists a world which humans cannot conceive of, one which lays beyond our level of understanding. As exemplified by the MIT example above, the computational power of AI is providing tantalising glimpses of that world. And when this power is combined with the creativity and ingenuity of humans, who knows what fabulous places it will take us.?
AI AUGMENTING HUMAN ENDEAVOUR IS THE FUTURE
Yes, of course AI can be used for great evil as well as great good, and of course there need to be boundaries and governance. But if you believe in the basic good of human nature and if you agree, as I do, with the likes of Steven Pinker, Hans Rosling and Bill Gates, that we are living in the best time of human history, then it’s not hard to see how the application of AI will usher in the next wave of the human race’s advancement.
Everywhere you look, AI is solving previously intractable problems. It has already made big breakthroughs in climate modelling, genomics, personalised medicine, astronomy and space exploration, and we’re merely scratching the surface of its capabilities. New applications are launched daily; at DAIVID AI is helping us use the world’s media to tackle the mental health pandemic by improving the world’s positivity ratio: upping the positive emotions being felt by humans while reducing the negative.
Quite where AI will take us remains a mystery, but I firmly believe that AI will become an augmenter of human capability rather than a replacer of humans or a destroyer of our world. For centuries, latent human potential and ingenuity has been unleashed by advances in technology - what would Beethoven have been without the invention of the piano, Marie Curie without the establishment of physics or chemistry as academic pursuits, or Pele without the introduction of football to Brazil? As AI enables a new generation of human geniuses, I, for one, can’t wait to see where this wondrous tech will take us.
Keep up to date with us on our LinkedIn page DAIVID ?? . Plenty on AI, #Emotions and of course #CreativeEffectiveness
★YouTube Guru★Video Marketing Expert★President and Co-founder of SEO-PR, an Award-Winning Content Marketing Agency
1 年Ian, what do you think about the current crop of AI video generators? Are they useful tools or are they just new ways to create mediocre content faster and cheaper?