Generative AI Economic Value Can Reach $10.3 Trillion, Accenture Reports

We’re clearly at an inflection point in relation to the last decades’ worth of technological advancements, and Accenture penned their timely perspective. They released a trove of new insights on the state of technology and Generative AI, sizing the potential economic value at upwards of $8 trillion by 2038, and revealing their predictions for 2024 and beyond.

According to Accenture research, Generative AI has the potential to impact 44% of all working hours across industries in the U.S., enable productivity enhancements across 900 different types of jobs and create $10.3 trillion in global economic value over the next 14 years. The estimation is a simulated GDP growth across 22 countries, where $10.3 trillion in economic value can be created by adopting responsible, people-centric approaches to Generative AI.

95% of workers acknowledge the value of Generative AI, yet only 5% of organizations are actively reskilling their workforce. This gap echoes with Databricks and MIT Technology Insights Review latest survey, where they found 40% of organizations’ struggles stem from workforce upskilling and training.

The advent of Generative AI is significantly transforming work processes and productivity. However, it also brings challenges related to job security, privacy, and ethical use.

Predictions in Generative AI, HMI, and Spatial Computing:

Technology Vision 2024 doesn’t reveal anything new to those who are mildly familiar with mainstream technologies. Accenture’s authors largely labor over contextualizing the most profound advancement we’ve witnessed in the last few decades; OpenAI’s ChatGPT, their underlying Large Language Models (LLMs), and the ensuing revolution. We did, however, find their predictions and timeline of key events leading up to present day, and what we might expect in the future, worth repeating.

Knowledge and AI Powered Information Retrieval

“With Generative AI, a digital butler is finally in the cards.” The report’s first section delineates a transformative shift in data interaction from a search-based to an advisor-driven model, facilitated by generative AI chatbots like ChatGPT. Knowledge sharing and retrieval are critical tasks for enterprises, and LLMs have the ability to serve as a universal interface.

Accenture’s timeline starts in 1665 when the first scientific journal was published. Fast forward a hundred years to the first edition of Encyclopedia Britannica published. Another 200 years forward and McGill University students created the first search engine, “Archie” and in 2001, Wikipedia was launched. StackOverflow came into play in 2008, when they started crowdsourcing Q&A for coding problems, leading up to the debut of ChatGPT fourteen years later. Here’s their predictions going forward:

1. By 2025, a leading airline will announce customer satisfaction with chatbot agents is on par with human agents.
2. By 2027, ML model dataset poisoning will be a top concern for enterprises.
3. By 2028, most enterprises will have their own chatbots to assist with knowledge management, research and task completion
4. By 2029, AI advisors will receive more traffic than search engines.
5. By 2031, a smartphone will launch where their app based interface is replaced with that of an AI agent based.

The AI Agent Ecosystem

Accenture foresees the emergence of AI agent ecosystems, predicting a shift in how enterprises strategize around intelligence and automation. AI will transition from assisting in tasks to taking independent actions, acting as interconnected agents that perform duties on our behalf. The success of such ecosystems hinges on access to real-time data, complex reasoning, and the ability to create tools autonomously.

Starting with the advent of the mechanical turk in 1776 – a tendentious concept still attached to today’s AI ecosystem. Fast forward to Apple releasing Siri in 2011 and eight years later, DeepMind’s Alphastar became a GrandMaster in StarCraft II. The advent of LLMs brought forward AutoGPT and BabyAGI last year, and their predictions are as follows:

1. By 2025, a new code repository will launch for open-source code written by agents.
2. By 2026, 75% of knowledge workers will use copilots everyday.
3. By 2030, 50% of mortgages will be approved and assessed by AI agents.

Human Machine Interaction

The “human interface” aims to understand human intentions for enhanced interaction with technology. Innovations like neurotech and brain-computer interfaces (EEG, fMRI and now fNIRS) decode neural data for communication, while devices like Apple’s Vision Pro allow gesture-based navigation. These advancements extend beyond healthcare to various commercial uses.

Accenture’s timeline starts in 1924 when the first EEG recording was made. Over the next 40 years, we witnessed the first instances of eye-movement tracking (to improve pilots performance), speech recognition (Bell Labs) and facial image recognition. The first BCI was implanted in a human in 1998 and Neuralink was founded ten years later. Last year, research partially return speech to a stroke survivor, and their predictions are as follows:

1. By 2026, a racing simulator featuring cars controlled by brain activity and eye movement will emerge
2. By 2027, a major retailer will launch a BCI pilot program for employee training and skill retention.
3. By 2029, a major bloc of nations will pass legislation protecting citizens’ neurorights.
4. By 2032, a large insurance provider will offer in-home gait analysis to detect early signs of Parkinson’s and ALS.
5. In 2035, we’ll be able to transcribe dreams with a neurotech device, transcribing them into visualizations and text.

Spatial Computing

Spatial Computing technology is set to fuse digital and physical realities. Despite skepticism marked by the “metaverse slump,” advancements continue. Central to this shift is the adoption of Universal Scene Description (USD) for creating 3D spaces and the exploration of sensory engagement, including touch, smell, and sound, to make digital environments more lifelike. “92% of executives agree their organization plans to create a competitive advantage leveraging spatial computing.”

Accenture’s timeline starts with the first multisensory immersive movie experience, Sensorama, in 1957. Xerox PARC released the graphical user interface (GUI) in 1975. In 1992, the first interactive AR system was created. Oculus VR was founded in 2012, Pokemon Go in 2016 and last year, Apple announced the development of a spatial computing system, Apple Vision Pro. Their predictions follow:

• By 2026, a professional sports league will launch an immersive 3D replay and highlight platform.
• By 2027, a major city will add spatial entertainment, directions, and information to public spaces.
• By 2028, a state public school system will announce offering physics courses taught entirely in an immersive spatial environment.
• By 2031, the gaming market will be dominated by VR and spatially immersive games.

Aside from providing an entertaining glimpse into the future, the report is a call to action for businesses to harness these emerging trends and foster collaboration.

The report includes input from an advisory board of industry veterans, interviews, and a substantial survey spanning consumers and C-level executives. The goal is to identify trends that will soon dominate C-level agendas.