More tech-led transformation is coming and artifical intelligence (AI) will be at the heart of that change, says a report by ARK Invest, which is an investment management firm that focuses on disruptive innovation in public equity markets.
"We're all about finding the next big thing. Those hewing to the benchmarks, which are backwards looking, are not about the future. They are about what has worked. We're all about what is going to work," says Cathie Wood, founder, CEO, and chief investment officer of ARK.
Also: 4 ways to help your organization overcome AI inertia
ARK publishes a Big Ideas report annually, which provides a comprehensive analysis of technological convergence and its potential to revolutionize industries and economies:
The ARK Big Ideas 2024 report is titled 'Disrupting the Norm, Defining the Future' and it highlights 15 big ideas, which are:
The 2024 Big Ideas report by ARK identifies five innovation platforms converging and defining the technological era: AI, public blockchains, energy storage, robotics, and multiomic sequencing. ARK also looks at the historic impact of converging technologies and estimated economic impacts.
The company identifies AI as the central technology catalyst. Let's take a deeper dive into the AI chapter of the 2024 BIg Ideas report.
The research suggests coding assistants are early success stories that have boosted the productivity and job satisfaction of software developers. AI-powered assistants are increasing the performance of knowledge workers and, interestingly, benefiting underperforming workers relatively more than high performers:
Foundational models are improving across domains. With larger training datasets and more parameters, GPT-4 outperforms GPT-3.5 significantly. Increasingly, foundation models are becoming "multimodal" -- supporting text, images, audio, and video -- and are not only more dynamic and user-friendly, but also more performant. The analysis says the cost of authoring the written word has collapsed:
AI-training performance is rapidly improving. What's more, training costs should continue to fall 75% per year. The research notes that after focusing initially on large language model (LLM) training-cost optimization, researchers are now prioritizing inference costs. Based on enterprise-scale use cases, inference costs seem to be falling at an annual rate of around 86%, which is even faster than the fall in training costs. Today, the inference costs associated with GPT-4 Turbo are lower than those for GPT-3 a year ago:
The open-source community is competing with private models. ARK states that the performance of open-source models is improving faster than closed-source models. LLM performance advances will require nuanced techniques, with the Stanford HELM framework highlighted as an example in the report.
Also: Why open-source generative AI models are still a step behind GPT-4
ARK also notes that computing power and high-quality training data appear to be the primary contributors to model performance. As models grow and require more training data, will a lack of fresh data cause model performance to plateau? Research institute Epoch AI estimates that high-quality language/data sources, such as books and scientific papers, could be exhausted by 2024, although a larger set of untapped vision data still exists:
The reserach suggests the acceleration of employee productivity using AI could have a multi-trillion dollar impact. AI has the potential to automate most tasks in knowledge-based professions by 2030, dramatically increasing the average worker's productivity. Software solutions that automate and accelerate knowledge work tasks should be prime beneficiaries:
The ARK Invest Big Idea 2024 report is the gold standard for detailing the disruptive technologies that will shape industries and economies for the next decade. The summary of the report's AI chapter includes potential possibilities for AI beyond 2030, and potential scenarios by 2040:
The report also includes other great content. The Digital Consumer and Digital Wallet chapters of the report are important for business leaders in any sector and any size. For example, the report suggests social commerce should grow 32% at an annual rate, from$730bn today to over$5 trillion in 2030.
Also: Five ways to use AI responsibly
The Big Ideas report is a comprehensive study of disruptive technologies, with more than 160 pages of analysis. To learn more about the research, you can visit here.