The AI revolution for businesses and the customer experience
After scouring the web and experimenting with numerous AI tools I am sharing my predictions for 2030 and beyond.
In this article I am outlining my 6 ‘I’s to how I see AI evolving over the next decade. They include AI’s impact on the Internet, Information, Innovation, Industry, Institutions and the Individual.
I spoke to one of my colleagues last month. Her son is studying computer science and he is wondering “what is the point of continuing if GPT-4 will write better code than me?”
Gen Z’ers are worried about their job prospects in fear of AI. In a study of 400 students at Hult Business School in 2019, 96% of students that they were worried about the changes AI could bring to their professions. And they have a reason to be even more concerned in 2023. Chat-GPT is one of the fastest growing apps in history. According to Similar Web Chat GPT had 100 million monthly active users two months after its launch in late 2022.
One of the key questions is how creative and ‘human’ will AI become?
AI will certainly get better at ideating, particularly when multiple AI tools are used together, but it will not replace human ideation and execution at a quality that will be satisfactory. Nor will it replace professional advice.
The Web will not die because of AI, and Apps will not be replaced by a chat interface. AI will need curated and validated Website data and content that is trustworthy, researched and validated by real humans. Yes, machine learning will actually improve the massive Large Language Model (LLMs) that exist, but it will still need Websites, Apps, and curated human data and structures that evolve from products and services. Therefore, Website structures, content, semantics, and schema will be key. We know this because AI will work in a similar way to how search engines like Google crawl the Web in tandem with LLMs.
The proliferation of AI APIs will create new value chains across different verticals and services. We will see new paradigms; some of which even the brightest futurists haven’t thought of because technology always surprises us. We will see new shifts and opportunities to make life better and create better services that are more accessible, entertaining, and innovative.
Where does ’big tech’ (Google and Meta) sit in all of this? I won’t outline the winners or losers in this article, but people like Zuckerberg and existing channels like social media will still matter and will still be around, just innovating in a different way. We don’t know (yet) if Google Bard will win against GPT-4, but this does not matter. Having a few big players put unprecedented resources into better AI will only benefit us, the digital community. What we need to do is ensure our brands and content are authentic and as invested as they can be.
Credibility in the AI era is everything. Similar to how Google’s Page Rank algorithm works, AI bots will be able to spot credibility a mile away so your corporate Website and personal online brand will be more important than ever.
Internet ‘2030’
The internet and the worldwide web have come a long way since 1996. We saw Web 2.0 in action but Web3 never materialised. AI is different. And it is not another crypto scam. Artificial intelligence is as “revolutionary as mobile phones and the Internet”, said Bill Gates on his blog.
Web 3.0 (not to be confused with the Blockchain, crypto and the Metaverse) will standardise information storage and data structure, creating a more intelligent Web. According to the World Economic Forum and Tim Berners Lee (father of the WWW) “Information would be stored in ‘database like environments’ called Solid Pods, which would be owned by individual users. While this is a more centralized approach than Web3’s use of blockchain, it would allow data to be changed more quickly because it wouldn’t be distributed over multiple places.”
This will be revolutionary as it will give users empowerment over the data ‘leasing it back’ to both big tech and future AI services. In this new Web, search engines will not be replaced. People will still need to search for specific things and will require a more curated brand experience, require rich media, and benefit-curated journeys where serendipity and inspiration is required. This is not to say that AI could not be embedded within branded experiences like Websites and Apps, for example.
Social media and advertising will be transformed too. Meta, the big tech conglomerate that owns Facebook, WhatsApp, Instagram and other products and services, are investing in technology that will massively help marketers with campaigns where an AI bot will be able to create rapid personalised campaigns to different market segments.
So instead of marketers using a single visual or infographic for a campaign, they would simply ask the bot to generate multiple visuals or graphics to work for different audiences. This wouldn’t replace designers either but will allow the designer to art direct and enhance what the AI created, as it would still need a human eye to verify the designs.
Information
Data mapping and data standards will be key. Websites and services that are opaque in how they extract data will not be trusted and regulation and laws (like Section 230) will change in due course. Furthermore, new regulatory bodies will be set up by governments (like the UK’s pro-innovation approach to AI regulation) to govern what data goes into AI and what data or utility is generated as we saw last week when Italy banned GPT-4 (blocked by the Italian data protection regulator, called the Garante).
Tim Berners-Lee (and his concept of Web 3.0) has been using Solid which is a new concept to ‘owning’ personal data in an editable and shareable in something called a ‘Pod’, which is a container for data. The concept is that any data a user generates on Netflix or Facebook will be stored by a user locally and then authenticated back to the vendor or brand.
This concept liberates users and will stop a situation where a third party controls your data. Berners-Lee uses Solid to organise and control the information in his personal life and store his bank statements, documents, photos, music, IoT data and exercise data on a Solid storage service on his personal computer.
In order for this to work in the real world, brands like Spotify, Amazon, Netflix and of course Social Media platforms and even governments will need to support these Solid Pods. Whether this will happen or not is yet to be seen, but the concept is logical.
Innovation
In another interview to CNBC Tim Berners-Lee said “When people look you up on Facebook, you don’t control actually what they see … Mark Zuckerberg’s algorithms control what news gets fed to them as they’re looking at your stuff”. How does this relate to AI?
Tim believes that AI technology will be so ubiquitous that we will have our own personal assistants scanning our data, helping us to learn, shop and study based on our own personal data.
These assistants will know us better than Chat GPT-4 and will be trusted as we will not be relegating our data to third parties. Some thought leaders are predicting that with the advances in computing, this AI technology will work locally not only on our own computers and phones, but also within IoT devices. Hardware companies like Apple and others may enter this space in due course.
In more recent innovations, we have seen examples of AI innovation that promises to write code, create computer games and create interactive 3D games.
My personal belief is that AI will never replace professional software developers, videographers and photographers but will help to create a new generation of ‘AI Prompt Producers’. The ‘professionals’ will always be needed to push quality and the devil in detail.
Some of the innovation will trickle into our work quickly, but other advancements may happen in years from now.
In the meantime, that are plenty of AI tools we can use today to push productivity and innovation forward.
I created a list of AI tools that I think can benefit businesses by copying some of the ones I saw in various social media posts.
Of course, it is important to check the various security and copyright polices of these (or any AI) tools before you use them and take note of the risks of using LLM tools, as outlined by the National Cyber Security Centre.
Industry 4.0
The third Industrial Revolution was the advancement of communication and digital. I see AI being part of the 4th Industrial Revolution as AI cannot work in silos.
AI will be supported and bolstered by 3D printing, Robotics, Data Science, hyper connectivity, quantum computing and of course new legislation and governing bodies. Each industry will be transformed in different ways by these developments. Organisations that adapt their structures and partners will be the ones that advance the fastest.
According to IBM’s Website “Industry 4.0 is revolutionizing the way companies manufacture, improve and distribute their products”
Organisations going through complex change and will still require service designers, design thinkers, data sciences, user researchers and data architects to unravel the intricacies of goals, complex relationships, workflows, processes and system.
According to Bill Gates’s AI essay last month (and a series of tweets from Lorenzo Green covering his essay), AI will actually aid help many professionals to do their job better.
According to the BBC, a report from Goldman Sachs said that AI could potentially replace around 300 million full-time jobs. Personally I strongly disagree with this analysis. Let’s look at history:
A classic example are phone operators.
As the number of telephones in the America multiplied at the beginning of the 20th century, so did the demand for telephone operators.
According to History.com in 1910, there were 88,000 telephone operators in the United States. By 1920, there were 178,000, and by 1930, 235,000 but technology advanced...
By 2020, there were 1.4 million telecom and ISP jobs in the U.S. If technology doesn’t create enormous amount of jobs then what does?
Health
As mentioned, in the future that Berners-Lee envisages that users will have all sorts of data stored in their pods — from fitness information to even health and electronic health records (EHR). After all, who should own your health records? But if Tim is right and you will own your data, then AI could use all that data to learn and be able to assist a user with insights, reminders and even preventative healthcare information.
GPT-4 and AI (or other new AI/ML tools) might eventually be able to extend their reach into third-party applications through APIs. For example, a diabetes monitoring app, EHR systems, and other health monitoring apps.
AI would be leveraged further with new third-party apps, which might it allow a user to to converting complex EHR data (such as a blood test result in a PDF) into a patient-friendly instant message. This would make it easier to understand medical terminology and of course result in much speedier communication when someone is sick. Joongheum (”PJ”) Park MD, an associate at Harvard Medical School, wrote a very good blog article about this: Unraveling the Role of GPT4 in Healthcare for Clinicians.
The impact of this technology is far reaching. Most people in developing countries do not have the privilege of even seeing doctors. According to WHO, half the world lacks access to essential health services. Even in the UK, one of the most advanced countries in the world, in April 2023 there are constant doctors strikes so seeing a medical can be difficult!
With more people in emerging economies (and even advanced economies) having access to smartphones that are perfectly capable of providing an AI healthcare function, this could be a game changer and might be able to provide better healthcare services and even save lives.
Beyond EHR, I can see AI making a big impact in TeleHealth and TeleMedicine helping users and health practitioners who aren’t doctors to better understand conditions and preventions. Of course, AI will not replace doctors and as of today, AI is renowned for making errors and ‘AI hallucinations’. So users should approach any AI generated medical ‘advice’ with much caution.
Education
The current education system is broken, particularly when it comes to nonneurotypical children and students. In the real world (or world of work) we are not assessed by our essays, university degree result or the quality of our code but by product quality, client satisfaction, and how we manage ourselves and others.
Take a software developer who is currently at university: Once this graduate gets a job he or she will rarely write code all day. Very few do. Programmers need to work with technology architects, client services, designers, and users, sitting in meetings (many meetings!). Very few universities can teach software developers real-world work skills, and AI will certainly not replace these capabilities. However, it can support developers to write better code and automate key parts of their work.
So a computer science lecturers will need to introduce AI into their curricula and ensure that using AI tools like GPT-4 are part of the syllabus.
Anderson Horowitz said that “AI is hardly the end of education”. Just as quickly as students started passing off the chatbot’s work as their own, new programs popped up to detect AI-written work, and teachers, looking to get ahead of their students, started integrating ChatGPT responses into their lesson planning.
Horowitz also said that with AI, it will become possible to personalise the learning modality (e.g. a visual learner versus an audio or text learner) and even personalise the presentation layer (e.g. get your favourite celebrity to present the material in a voice you love). Let’s wait and see ;)
AI might also be able to detect a student’s competency and skill levels and fill the voids by getting the software to test and track your knowledge and proficiency. This could help the learner to fill the literacy gaps faster, thus leading to much more interaction between the learner and content.
Finance
Bank statements and banking services could be more accessible through powerful AI bots. Imagine not having to scroll through long statements to find information in a PDF and how easy it would be check your mortgage eligibility through a simple text interface.
According to the FT and Jeremy Mugridge of Quilter (a wealth management firm), in prior years, Google search was a game changer when it came down to users learning about financial products. New research from Quilter stated Chat GPT can take personalised banking advice to the next level and that there is ‘huge interest from the public’.
The public interest stems from AI’s potential ability to generate sophisticated responses to a wide range of financial requests based on their individual situations. With regard to loans, financial institutions would however need to be more transparent on eligibility criteria and would need provide an open and transparent scale, particularly when it comes to underwriting and other decisions.
In terms of functional utility, chatbots should be able to provide balance enquiries and transfer funds reducing taps and clicks through typing or even voice user interfaces. This would reduce the workload from call centres and other channels and help to free up the bank staff.
Besides having more transparent business operating principles, banks would need to create streamlined data structures and better information repositories on their websites. New technology like AI would require banks to study and remap user journeys to ensure that even through a chat interface there is a logical flow for the utility to fulfil the user need from start to finish whilst ensuring there is a feedback-loop, escalation points and of course, security and privacy.
An AI bot interface (a powerful like GPT-4) could dramitcally simplify accessibility and utility. In the UK it may fulfil some of the requirements of the Financial Conduct Authority (FCA) recently announced Consumer Duty Act. Among them, it Consumer Duty states that consumers should be given the information they need, at the right time, and presented in a way they can understand.
See an AI hip-hop video about Consumer Duty and Service Design I generated using a few AI tools in less than 45 minutes!
Of course AI will not replace IFAs, bank staff and human interaction but AI can be very empowering for consumers, if they obtain faster access to transactional banking functions, tools, data and information in a simpler language that they can understand.
Institutions
As we saw in the innovation section, institutions can use a myriad of AI tools. The two big questions is will these tools make human roles like developers, managers and designers redundant in the next few decades. If the answer is no, how will these tools impact our productivity and will this AI create new professions? I think the answer is yes. Some of the professions that I see being created are Prompt Engineers, AI Ethic Supervisors, AI copyright specialists to UX AI Researchers who leverage real user input (such as qualitative interviews) with new AI tools.
I saw a recent post from Heather O’Neill a Service Designer based in NYC. Here is an extract:
“There’s now an AI tool that claims to be able to give user research insights without talking to users. First of all, that’s called guessing, even if it’s “educated” guessing. Second, by inputting your research data into ChatGPT, AI tools like that one can leverage that data over and over again. Ooops. ChatGPT and similar AI tools are fun. They’re powerful. They can help us with a lot. But we need to be thoughtful and cautious about how and what we share and make use of them”
“The worry is not so much that ChatGPT will lead to large-scale unemployment — but that companies will replace relatively well-paying white-collar jobs with this new form of (AI) automation, sending those workers off to lower-paying service employment while the few who are best able to exploit the new technology reap all the benefits.”
I agree with O’Neill wholeheartedly. Fake AI generated user research is dangerous to product teams as it will generate the wrong outputs.
In this Next Web article, The World Economic Forum predicts that the technology will create 97 million new jobs by 2025. “Jobs specifically related to the development and maintenance of AI and automation will see growing adoption as AI integrates across multiple industries”
Just look at your business today. Did a business like yours have 20 designers and 38 software web developers 25 years ago? No as these jobs did not exist. Did your business even exist 45 years ago? For many the answer is yes but it is most likely that how you ship and provide your products and services has fundamentally changed and you rely on all sorts of ‘new professions’.
Efficiency will be key and the future is closer than you think. My colleague Chris Carrington-Hughes, a Project Manager at Cyber-Duck wrote about how AI can automate meeting actions and support with resourcing freeing up bandwidth and efforts.
Individual
The AI chat interface is already proving valuable support for visually impaired users (the BeMyEyes App) who used to rely on fellow humans to help them identify objects through a video call. In a BeMyEyes survey (featured on Mashable) polling low-vision users, the main feedback on obstacles to getting fellow people to detect objects was that some users actually felt uncomfortable with other people seeing their intimate spaces and intruding on their privacy.
The AI version of the app (in limited beta) can now use Chat GPT-4 to recognise images. A visually impaired user can request a real human help if they don’t feel comfortable with the AI assistant or the result.
My Cyber-Duck colleague James Martin wrote about how AI will enhance work like thematic analysis, but certainly not run or facilitate user interviews. Users will still want to talk to individuals, and humans will be able to connect better with other humans on a qualitative research level than a machine. In the foreseeable future, even the most advanced artificial general intelligence (AGI) will not be able to build an intimate relationship with another human. The AI will also be unable to fulfil the complexities and the logistics of organising and facilitating interactions between the researcher and a myriad of users who have a variety of complex life situations. This requires a strong human touch after all.
People who rely on AI too much (e.g. by jumping straight into the AI bot interface to ideate) will lose out to professionals who still jot things down to plan things in a more abstract fashion away from tooling. A recent comparison of using a paper notebook versus typing into a digital devices, used Functional magnetic resonance imaging (fMRI) to identify specific brain activation regions differences during memory retrieval.
According to Psychology Today, the study in 2021 highlighted:
“…that handwriting in a notebook triggers more robust brain activity. Writing by hand is associated with stronger neural encoding and memory retrieval”.
This study highlights that people who really want to progress in their careers and master their domain must still use analogue ways of thinking, working, have a process to it, leveraging AI tools with caution. After all, you can’t become an ‘authority’ and have confidence without deep thinking. It simply cannot be outsourced to an AI tool!
Conclusion
AI brings the promise of revolutionary new applications, but it will not replace human creativity. AI will provide a suite of tools to unleash progress in a wide variety of fields. It should also be used with caution due to its susceptibility for misinformation and manipulation as well as high jacking our data and infringing copyright.
The creation of Internet is a great example that sets a precedent for AI.
With the Internet we collectively developed the world’s most advance communication and productivity network in history creating millions of new jobs and new professions.
One of the Internet’s by-products is the World Wide Web. It gradually became more and more advance, creating thousands of templates, code snippets, design widgets and business plan templates. This did not remove jobs. It created more jobs. The Web simply made us more productive, faster and inspired. This is why I think AI is an enabler not a destroyer. It will inspire the business community and nudge us to upskill and evolve our roles.
Yes, AI will be able to create business plans, keynotes, videos, and video games, but we have had tools that can do this before. Yes, the new generation of AI will be better than the previous tools, but it is naive to say that these new AI-generated results will be product market fit. Human intervention and deep work will be required.
Trust in non-branded content and or sites that host user generated content will drop and the Tim Ferris effect will compound in the age of AI where real people will be trusted more than ever.
In a bias-filled world, where AI provides factually incorrect information or news, checking the facts will become more important than ever. Some people are fooled into believing everything that AI generates. Will you be one of these people?
A University of Washington report featured in the Wall Street Journal highlights that more than 70% of people reading an AI generated article thought the content was correct and true, despite the facts being wrong!
So to my colleague’s son at university, I say this:
Have faith. AI won’t replace you — use it to your advantage, but don’t forget to stimulate your brain ‘analogue style’ and be creative!
