Why Artificial Intelligence (AI) will be the technology of 2023 and beyond
A look at its past, present and future of AI of 2023 and beyond
Only 10 years ago, barely any machine could reliably provide language or image recognition. Today, machines have learned to outperform humans on many tasks. In the past few months, we’ve seen progress in AI capabilities that has impressed even skeptics. A “golden decade,” one researcher called it. In 2023 and beyond, we will see more such systems (especially Generative AI systems like ChatGPT) complementing or replacing us human creators in many areas.
Every year there is a new technological achievement: blockchain, 3D printing, Web 3.0, and the metaverse. So, what is the technology of 2023?
Artificial intelligence (AI). Even though I’ve worked extensively with this technology for 10 years, we are amid a significant leap forward in artificial intelligence. Just in the last few months, we’ve seen advances in AI capabilities that have impressed even skeptics. But first, let’s take a few more steps back and begin by looking at AI’s development in the past decade.
Index
2012-2014 – The beginnings of image recognition, reading comprehension, and language understanding
Some researchers say that the year 2012 was a milestone for deep learning. It was the year that Google researchers built a large neural network with 16,000 processors and a billion connections to recognise images and videos of cats. This is an example of reinforcement learning, which, along with supervised learning and probabilistic program induction, has been one of the most successful AI frameworks of the last decade. Recognising pictures of cats may seem like a small achievement. But at the time, machines were just starting to use deep learning for image recognition. In 2012, image recognition was still in its infancy, and in tests comparing AI to human performance, AI was found to perform at around -40, which was still below human performance (here set at the zero base line). In addition to image recognition, AI also underperformed humans on other tasks a decade ago, including reading comprehension and speech understanding. Despite the invention of NELL (Never-Ending Language Learning), a semantic machine learning system, in 2013, AI was still unable to reliably perform language processing tasks. AI’s ability to recognise language improved dramatically with the invention of Alexa in 2014. Before that, Apple’s Siri allowed users to manage their phones using speech. However, AI’s ability to understand speech was still inferior to that of humans. In the following years, AI reached a level of speech understanding that was better than that of humans. This was due to improvements in AI speech recognition, advances in language processing and related neural network language models, and information organisation. Although AI systems still struggle to produce long and coherent texts, chatbots such as ‘ChatGPT’ show the immense progress that has been made to date.
2015 – 2017 AI began performing better than human beings
2015 was the year when everyone was allowed to build meaningful AI models. After IBM’s flagship artificial intelligence system, Watson, became famous for beating Jeopardy! champion Ken Jennings in 2011, several open-source platforms for machine learning came to market (such as Google’s open-source deep learning framework, TensorFlow). This has enabled companies and developers to work with the technology in new ways. The year also saw significant advances in face and image recognition. For example, machines beat humans in the 6th edition of the ImageNet Large Scale Visual Recognition Challenge (ImageNet is a standardised collection of millions of photographs used to train and test visual identification programs). Looking ahead, in 2016, deep reinforcement learning – a combination of neural networks and reinforcement learning – generated massive hype in the AI community when Google’s AlphaGo beat the world’s best Go player. Furthermore, in 2017, the use of self-supervised learning models combined with deep neural networks increased with the introduction of transformers. Today, these Transformer models are the mainstream approach for natural language processing (NLP), including applications such as machine translation and Google web search.
2018 – 2019 Data Security, Language Processing, and AI in Medicine
Thanks to the Cambridge Analytica scandal, 2018 was the year when the topic of data security came to the fore. In line with this, a McKinsey survey found that in 2018, risk was one of the functions where most respondents said the value of AI was visible. In addition, language processing took a big leap forward in 2018 with the development of BERT. BERT is an example of a neural network language model that learns about word usage, grammar, meaning and basic facts in different contexts. By connecting sequences of words simultaneously, rather than stringing them together from left to right, models like BERT can produce summaries that are almost indistinguishable from human-generated text. These language processing models are crucial for supporting applications like chatbots, and have driven their immense progress over the past decade. Furthermore, in 2019, researchers began to create an AI system that outperforms human radiologists in detecting lung cancer. This was achieved using a deep learning algorithm that can interpret computed tomography (CT) scans to predict the likelihood of someone having the disease.
2020 – 2021 Quick AI advancements due to the pandemic
In 2020, AI development was driven and accelerated by the COVID-19 pandemic. AI was largely responsible for speeding up vaccine development, which would normally have taken several decades. Instead, this process was significantly shortened because AI helped researchers analyse huge amounts of data. The growth of AI is exemplified by global corporate investment of $68 billion, an increase of 40% from 2019 to 2020. Furthermore, in 2021 alone, the number of patent applications related to AI innovations will be 30 times higher than in 2015, demonstrating the rapid progress of AI development. Over the past year, the research community has focused primarily on the application of AI to computers. This subfield teaches machines to understand images and other visual material to perform well in image classification, object recognition, mapping the position and movement of human body joints, and face recognition.
Today – AI has become indispensable in our lives.
Artificial intelligence (AI) has developed rapidly over the past decade. Just 10 years ago, few machines could reliably recognise speech or images. Today, machines have learned to outperform humans at many tasks. For example, AI systems can detect fraudulent charges before you know you’ve lost your card, or check eligibility criteria when someone applies for a loan. AI recognises patterns and evaluates options in our daily lives. It knows which Instagram post I like and keeps me on the social media platform, which price on Amazon makes me want to buy, and whether I should leave my AirPods at home. In recent months, there has been a particular explosion in ‘generative AI’ – systems that create new possibilities. We’ve seen advances in AI capabilities that have impressed even sceptics.
A firework of Generative Artificial Intelligence
Within a few months of one another, three different image-generating AIs were made public: Dall-E, Midjourney, and Stable Diffusion. You enter a text, and the system generates an image within a few seconds. Ask for “an astronaut riding a horse on Mars,” and the AI gets started. One image of Midjourney was on the cover of the Economist in June 2022, and another won an award at the Colorado State Fair. A sign of what’s to come in 2023.
For me, the developments in text-generating AI are even more exciting. The American startup OpenAI has created a worldwide hype with its chatbot “ChatGPT”. This chatbot mimics the neural network of the human brain. The bot can hold human-like conversations and generate convincing answers to even the most complex technical questions and dialogues. This is a major milestone for AI systems, as language processing has proven to be one of the most challenging tasks for AI in the past.
AI adoption in companies is on the rise.
This year’s State of AI report shows the accelerating progress also in the industry. In 2020, not a single drug was in clinical trials that had been developed using an AI-first approach. Today, there are 18. An AI system from BioNTech successfully identified numerous high-risk covid variants months before WHO’s tracking system detected them.
According to McKinsey, AI adoption has more than doubled since 2017. Robotic automation and computer vision are the AI capabilities most frequently deployed each year. Moreover, investment in AI has increased. Today 52% of respondents report that 5% of their digital budget is used for AI investments. In 2017, that proportion of respondents was only 40%.
However, AI adoption is still concentrated among the AI high performers. This means that the companies that have led the way in AI are still building their competitive advantage today. These AI leaders are more engaged in the ‘industrialisation of AI’, linking their AI strategy to their core business practices. The reasons why these companies continue to outperform are that they invest more and spend more than AI laggards, which then attracts more and better (tech) talent.
What is to come for AI in the future?
By 2023, we will see more such systems complementing or replacing us human creators in many areas. Video generation is under development, as is the creation of customised media. (“Siri, show me a 90-minute film about a CEO who gives up her career to make her fortune as an actress, Steven Spielberg style”).
But while there has been tremendous progress, and more companies have adopted AI, there is still a lot of research to be done before we can produce fully general-purpose AI systems. Over the next few years, we will see further advances in self-supervised learning models, continuous learning and task generalisation. For example, future applications of neural network language models such as BERT could enable human interaction across languages and contexts. We are likely to get higher quality data for languages used by smaller populations, and biases in AI systems will be easier to detect and remove. We can therefore expect to see AI chatbots such as ‘ChatGPT’ become more human-like and more widely used, for example in schools and universities.
In the coming years, we will see an increased use of facial recognition for access control and security. China is one of the most prominent examples where the government has almost fully embedded facial recognition technology in its society. However, it remains to be seen whether more countries will follow China’s lead, as this ubiquity of AI in society is likely to raise privacy, surveillance and ethical concerns.
What is certain is that the adoption of and investment in artificial intelligence will continue to grow in the coming years. In fact, according to the McKinsey survey, 63% of respondents say their AI investments will increase over the next few years, generating benefits such as positive revenue effects and reduced costs. By 2025, global revenues from AI-related enterprise applications are expected to reach USD 31 billion. However, the growth of AI must be transparent and managed to address privacy and ethical concerns. Governments and businesses must therefore create AI laws today that will guide AI development in the future.
Comments are closed.