AI Predictions 2023: The Year AI Stole Every Household’s Imagination

Dr. Santanu Bhattacharya
8 min readMar 28, 2023

--

Image Credit: Cottonbro Studio from Pexel

Decades from now, the two years of the post-COVID world 2021 and 2022, will be seen as watershed moments in Artificial Intelligence (AI). In the unprecedented times following a pandemic-ravaged world, enterprises accelerated their journey to data driven digital transformation by finding ways to creatively, boldly and broadly apply AI to survive in the short term, while also thrive in the long run.

A clear pattern has emerged globally for the past 5 years in terms of development and adoption of AI. In 2018–19, enterprises started their AI experimentations and created a “recipe” for future development and a roadmap for adoption. In 2020, adoptions began in an earnest way and suddenly, COVID-19 gave the business leaders an opportunity and impetus to push automation and AI. In 2021, the fallout from the second and third waves of COVID-19 became clear, forcing rapid decline of many traditional, not-so-digital businesses.

In the USA alone, brands that lasted for decades either declared bankruptcy or shut down for ever: JCPenney, Lord and Taylor, Furniture Factory Outlet, Hertz, Sizzler, and not so lucky anymore, the Lucky Brand. In 2022, finally out out the shadow of the COVID, human-focused AI applications proliferated.

As the C-suite has now taken notice, following are the relevant trends I expect to emerge in 2023

AI Will Become More Human-like in Nature

I start by my 2023 predictions with the same headline as last year, which read as:

“As AI technology advances, it is becoming more human-like in its interactions. This means that AI systems will be able to understand and interpret natural language, making them more intuitive and user-friendly”.

They came out eerily accurate leading to the biggest AI story of the decades, the huge growth of “human like” AI, namely, ChatGPT. While other AI consumer-facing application such as Dall-E have existed since 2021, ChatGPT was the first one to go viral and catch the imagination of the wider public for its ability to produce answers that are human-like.

In 2023, enterprises will rapidly prototype and start deploying ChatGPT based customer interaction bots — chat and voice. Some will even start integrating these capabilities with AI systems that will be able to recognize and interpret facial expressions and emotions, allowing them to respond appropriately to different situations through video chat conversations.

GPT-4 Will Start Changing Human Work the Way We Have Never Imagined

My own prediction for 2022 will come true a bit late. My prediction for 2022 for GPT development was:

“Expect this hockey-stick growth to continue in 2022 . I am personally rooting for GPT-4 with 100 trillion parameters”

With GPT-4 release being imminent in 2023, the impact on work we do as a human being will start changing noticeably. While GPT-4’s capabilities are yet unseen to most, it can potentially

  • Enable full feature length movie script creation, automated video and audio generation — think of an entire movie created, automatically, in a day — pouting movie makers at stake, or letting million of yet undiscovered movie makers flourish
  • Create personalized coaching based on one’s career progression, life’s history, goals — “democratizing” the professional coaches industry which today, can be accessed by a selected few in the top of their professional careers
  • Provide legal advice at the level of a mid-level lawyer. While it can not yet represent a client as most countries bar such actions, in less than 5 years, a small European country or a state in the USA would likely allow such representation, perhaps in an assistive role
  • (While it has been postulated for some time), my personal bet is that in the 5 -7 years timeframe, by the time of GPT-8 or GPT-10, low-to-mid level software developers job, as envisioned today, would be done by a machine, truly embarking into the world of Software 2.0 (put a like for software 2.0), where “software writes the software”
Figure 1: Large Language Models Such as GPTs will change the way humans work significantly. Image Credit: Matheus Bertelli from Pexels

Multimodality Will Find Unique Use Cases

Multimodality refers to the way we humans experience through exposure to sensory cues. A movie is experienced through hearing and viewing, and in the case of 4DX movies, also includes the sense of touch — for example, air blowing or water spraying on your body etc.

Most AI development to date has been single mode — text, numbers, image, videos etc. With the Large Language Models (LLM) and GPTs making rapid strides, multi modalities would open up many applications that we have not imagined possible with AI before.

The end-to-end movie creation is an example of a Multimodal AI. Multimodal experiences can be especially useful in healthcare AI — where diagnosis can be made from multimodal analysis of a patient’s description of a symptom, analysis of tests (blood, MRI, other) and previous treatment data. The day of Star Treks’ holographic doctor who can do a human level diagnosis may not be far away

Business Operations Will Likely Be the Largest Beneficiary of AI in 2023

While AI has suddenly become mainstream due to rapid growth of ChatGPT, AI is most impactful in business operations that are closest to the consumers and, therefore, create the highest values. Examples include service operations such as contact centers, product offerings and marketing.

While enterprises have been automating their contact centers and to a lesser extent, marketing using AI, the biggest segment of beneficiaries this year would be the small business. Consider this scenario: a niche snack brand that does a few million dollars in revenue in a regional US market.

For a new product launch, pre-2020, they would spend $250–500K with a marketing agency for concept generation, testing, developing packaging and marketing campaigns, not including the cost of the media.

Today my estimate is that the same can be done for about a tenth of the cost if AI tools are used extensively. In fact, to test that out, we have created product packaging concepts for “free”, something that cost a minimum of $75K from a US based media agency.

Figure 2: Packaging design for a new product launch, which used to cost hundreds of thousands of Dollars and take months, now can be done in days with minimal costs. Image Credit: Stable Diffusion

Personalized Recommendations Will Happen Before You Know You Need Something

An interesting report from Accenture claims that 91% of the consumers are likely to shop with brands who send them “relevant” information. Also while Generative AIs are claiming the headlines, we expect many existing products will continued to be built on improving the existing recommendation systems.

While the content providers such as Instagram, or Spotify have steadily improved their offering and consumer engagement with better recommendation, the ability to generate “synthetic” content rapidly can upend the field upside down, where recommendations are co-created with contents or products customers are likely to engage with or buy.

2023 will see a rapid scaling of personalization AIs across the spectrum, and they will become the “new normal” of our online and in-store shopping and digital product experiences.

Search Will Change the Most in 2023 Than it Did in the Past Decade

Search is the heart of the modern internet, it is the way we find and navigate digital information and experience, be it on Google, Amazon, or Netflix. The way today’s LLM can read, summarize, and write information, was inconceivable just two years back.

ChatGPT has suddenly opened up new ways to conduct conversational searches. It is a revolutionary AI technology that has opened up diverse avenues. Why type search texts, try best to guess which keyword will give you the best results and then navigate through many links for creating a trip itinerary to Vietnam, when you can simply type “create an itinerary for a 5 day trip to Vietnam” and get a customized answer within a minute.

While LLMs will not kill Search the way we know it today, it will definitely undergo one of the most dramatic redesigns since the design of modern search in the early 2000s.

AI Ethics and Regulation Will Become More Important

As AI becomes more advanced and ubiquitous, concerns about its ethical implications and potential risks will increase. In 2023, we can expect to see greater scrutiny and regulation of AI technology, particularly in areas such as privacy, bias, and accountability. Organizations that develop and use AI will need to be transparent about how their systems work and ensure that they are not perpetuating or amplifying existing social inequalities. Significant research is now available on the ethical implications of AI, and businesses will rapidly evolve in 2023 to meet these challenges.

AI is poised to continue transforming many industries and creating new opportunities in the years to come. However, it is important to consider the ethical and regulatory implications of AI and ensure that it is used in a responsible and transparent way. By doing so, we can harness the full potential of AI while minimizing its risks and drawbacks.

Figure 3: The ‘goodness’ delivered by the growth of AI Applications has to balance against the need for trust, privacy, and digital rights. Image Credit: Mati Mango from Pexels

The Balancing Act: AI With Trust, Security and Digital Rights

In 2023, businesses and governments across the world, finally out of COVID’s shadow and employing AI in innovative ways to deliver better governance to their stakeholders — customers, business partners, citizens and more.

The challenge will be to do this at the right pace while building the stakeholders’ trusts by protecting their rights and their digital security. Those who manage to do so will win the heart and soul of their constituents.

The Global AI Action Alliance, a group of World’s top 100 companies, institutions and think tanks under the aegis of World Economic Forum, is an example of an effort to spearhead such conversations globally as AI sets to get deeply integrated to human societies around the world.

This year is set to become a pivotal year in the continued development and application of artificial intelligence. As AI systems become more sophisticated and widely adopted across industries, we can expect to see significant advancements that will bring a renaissance in areas such as healthcare, finance, transportation, and more. However, as we are still struggling to find a way-out for potential job displacements due to AI, it will be a delicate time to approach it with caution, foresight, and a commitment to ethical principles.

Previous Story: Prediction 2022: The Year AI Becomes Human Focused

Next Story: Sage Who Dreamed of the Atom

--

--

Dr. Santanu Bhattacharya
Dr. Santanu Bhattacharya

Written by Dr. Santanu Bhattacharya

Scientist at MIT Media Lab. Former Chief Technologist at NatWest, worked for NASA, Facebook & Airtel, built startups, and future settler for Mars & Tatooine

Responses (1)