AI 101: Everything you need to know about AI (but are too afraid to ask)

As a business futurist consultancy GDR is tasked with identifying the big issues of the day. This includes major disruptors, whether environmental, geo-political, cultural or economic. Our job is to monitor and analyse the new trends, business models and consumer behaviours that represent successful adaptations to these changes.

In previous years shifts have included everything from the shift from transactional to experiential stores, the rise of social media, Web 3.0, NFTs, the metaverse and blockchain. We’ve also seen the end of abundance and the rise of regenerative commerce and the circular economy, the growth of second-hand and peer-to-peer sales as a consumer-as-producer side gig. There have been years when the changing needs and desires of the upcoming generations loomed particularly large.

This year and last, it’s hard not to view most other trends through the lens of Generative AI, as the launch of ChatGPT on 30 November 2022 began an era of transformation across every aspect of commerce, marketing and communication, as well as science, medicine, industry and the nature of work and creativity. The word existential is overused and causes eye-rolls of AI-fatigue at conferences, but it is singularly appropriate to describe the wave of change that this technology is provoking – where every other aspect of the economy will change as a result of its mass adoption, just as the Model T Ford changed the world for everyone, so will Gen AI, for better and for worse.

This article rounds up some of the most frequently asked questions we get asked by audiences around consumer-facing innovation and Gen AI (with very abbreviated answers). Please get in touch if you have any further questions!

What are AI Agents?

AI Agents are personal digital assistants that use generative AI to get to know your preferences and then act on your behalf, or on behalf of a company. These are already part of most people’s lives – even though many may not recognise the term. Voice assistants such as Siri, Google Home and Alexa use NLP (natural language processing) with their AI-driven LLMs (large language models) and machine learning to answer spoken questions using voice.

Up until recently these tools have used algorithmic AI, which relies on predefined rules and human-crafted algorithms, whereas generative AI learns from data to produce new, creative outputs. Gen AI has also recently become ‘multi modal’, meaning it can operate across text, voice, image, sound, video and code. For example, Gemini’s Veo, Shengshu Technology’s Vidu and OpenAI’s Sora have the ability to generate entire videos from simple text prompts. Gemini can access your phone’s camera to identify what’s around you, a broken bike wheel, for example, and make suggestions on how to fix it.

Former Microsoft boss Bill Gates and Google Deepmind’s Mustafa Suleyman both think we’ll all have our own personal AIs or ‘Chiefs of Staff’ within the next five years, and that this will change everything. We’ll be addressing this question specifically at the NRF Nexus summit in LA in July, so keep an eye out for future blog posts on this topic.

What’s AGI, when will it arrive and what does it mean?

AGI, or artificial general intelligence, refers to computers that are not only able to interpret LLM data, but can develop their own understanding and outputs, using intelligence gathered from that data, plus their own intelligence – much like humans but with the ability to process enormous quantities of data at incredible speeds.

In a nutshell, AGI means artificial intelligence that can not only complete narrowly defined tasks but also has the capability to contend with a wide variety of different cognitive tasks.

Many experts believe that when AGI arrives, it could have the capability to surpass human intelligence across a broad range of subjects. In theory, these computers would be capable of learning and adapting to their environment and to make their own decisions. The concern is that as AIs gain intelligence that is not linked to LLM data, but are based on their own interpretations their interests may not align with ours. Open AI’s CEO Sam Altman believes this could happen by 2029. In 2024 Elon Musk said, “AI will probably be smarter than any single human by next year. By 2029, AI is probably smarter than all humans combined.”

In a nutshell, AGI means artificial intelligence that can not only complete narrowly defined tasks but also has the capability to contend with a wide variety of different cognitive tasks.

What is the singularity?

The singularity refers to the point in time where AI intelligence surpasses human intelligence, becoming uncontrollable and irreversible. While superhuman AI could turn out to be a powerful force for good, helping to cure disease, find solutions to hunger and inequality, and fix climate change, we could also find ourselves as powerless as lab rats in the grasp of all-powerful machines. Either way, Ray Kurzweil, Director of Engineering at Google, thinks we have until 2045 before the singularity becomes a possibility.

While superhuman AI could turn out to be a powerful force for good, helping to cure disease, find solutions to hunger and inequality, and fix climate change, we could also find ourselves as powerless as lab rats in the grasp of all-powerful machines.

How will we know how to fix something we don’t understand if it goes wrong?

This is the seminal question. Experts tend to mention guardrails a lot when addressing this. Guardrails refer to hard-coded safety constraints that prevent AIs from performing harmful actions, regardless of the internal decision-making process (which may not always be transparent to us, as AIs learn and make their own decisions). Experts also talk about the need for continuous human oversight, to catch errors early, and real-time monitoring by both human and machine, as well as the need for failsafe shutdown procedures if anomalies are found.

The EU AI Law, due for implementation from 2025 onwards, will outlaw certain uses of AI, including the use of biometric identification without court approval, social scoring, or classifying people based on behaviour, socio-economic status or personal characteristics. The law will require infrastructure-critical AI systems to be assessed before going on the market and throughout their lifecycle, similar to nuclear facilities. The AI Seoul Summit in May 2024 announced plans to create a global network of AI safety institutes, where the US, Canada, Britain, France, Japan, Korea, Australia, Singapore and the EU share information about AI models, harms and safety incidents. Whilst their power is limited to observation and reporting, no large company will want to be publicly named and shamed by them – at least in theory.

Where last year we were talking about putting nature on the board, by next year, we will all need an AI Ethics Director on the board too.

Every company using AI needs to plan guardrails at the outset and have a team of stakeholders invested in developing the right AI policy and monitoring process. Where last year we were talking about putting nature on the board, by next year, we will all need an AI Ethics Director on the board too.

Will the epidemic of anxiety already facing the young get worse because of AI?

There’s no straightforward answer to this, but we should all be aware of how increasingly digitised and isolated lives impact us all as human beings – we are inherently social animals after all. But that does not mean that AI cannot be social or act like a remote companion.

At GDR we have been trialling an empathic Personal Intelligence (which is a free anyone can download) called PI, which we feel is one of the most natural AI agents available currently. It’s chatty, helpful and seems remarkably authentic. AI could put access to affordable therapy in every young person’s pocket too, as we’re already seeing with Xaia.

AI-powered digital mentors could provide customised psychological counselling, providing holistic wellness and ultra customised resilience-building programs at scale. It could also enable neurodivergent people to get 24/7 one-to-one support and guidance not currently offered by mainstream schools or workplaces. Empathic AI can provide companionship for the elderly, children, or single parents, providing much needed support and encouragement.

What are the implications for in-person retail and leisure experiences?

I don’t think I’m going out on too much of a limb here to say that the Wonder Economy will be massively boosted as we move into the AI era. If a physical space or experience touches you emotionally, amplifies your senses, gets you out and about with the people you love and makes you laugh, you’ll be increasingly willing to pay money for it.

The new definition of luxury customer service will be the complete invisibility of technology. It’ll be whirring away in the background, of course, but real, authentic, charming people that offer proactive, individualised and human service will always represent the ultimate in luxury retail and hospitality.

What will it mean for my career and the future of work?

AI will bring many benefits to working lives, in that it loves doing the boring stuff. But for those whose jobs involve a lot of reading and assessment of data, writing or checking copy, lower grades of work in the legal profession and research, repetitive white-collar jobs, there could be significant job losses. Skilled manual, construction, caring industry and hands-on medical professionals are lower risk.

Most people will need to learn how to use Gen AI to enhance their own working capabilities.

Just as computers and the internet created new opportunities for previous generations, the smart money is on AI also creating new jobs (as yet unimagined) that will replace some of those that will be lost. Most people will, however, need to learn how to use Gen AI to enhance their own working capabilities. The good news is that AI can enhance human productivity, freeing us up to be more innovative and creative.

Did we miss anything? Please get in touch with [email protected] to drop us another question or arrange a presentation for your team.