Generative AI in business: A marathon, not a sprint

In just over six months of free access to the public, ChatGPT was able to turn the strategic technology plans of most companies worldwide upside down, regardless of sector or size. It is not that Artificial Intelligence is a new topic on the corporate agenda; in fact, it has consistently been among the top investment priorities in recent years. But then, what has caused the significant shift? What are companies looking for – or should they be – and what practical changes do they need to make to their previous plans?

OpenAI's tool is the most renowned representative of a new class of solutions and platforms and one that has been proliferating ever since its introduction. The fact that they provide access in a creative, massive, and, to a significant extent, democratic manner to the most recent advances in the area of Gen AI can be seen as some of the reasons for its widespread adoption.

The world discovered at the end of last year that the machine's ability to understand, process, and generate language, especially in text format, has advanced much faster than previously anticipated. Even Bill Gates, who has been following the OpenAI team closely since 2016, confessed that, when witnessing the advances of ChatGPT in the last year, he was faced with "the most important advance in technology, from the graphical interface."

For this reason, Yuval Harari, one of today's best-known and most influential historians, explained that "language is the operating system of human civilization," the tool from which virtually all of our culture, laws, religions, histories, and areas of knowledge are made. What ChatGPT and its ilk did was "hack" this operating system. Computers are now able to create language in different formats (text, audio, video, software, etc.), not only in virtually infinite quantity, as they were already capable of, but in equivalent quality or, in growing cases, better than the vast majority of humans.

If the probable ethical and social consequences of this new moment are still the subject of intense debate, with very polarized views and forecasts, the substantial economic impact for companies is already taken for granted. Most CEOs and investors have already realized that any human intellectual activity, in the corporate context, that depends on or makes use of language may soon be radically optimized and improved. This brings about fascinating possible developments, especially in two dimensions:

The first, most apparent, and immediate one is productivity and efficiency. Knowledge workers will soon have powerful "copilots" to help them. No matter what the profession, whatever our area of expertise, the "smart and global interns," as defined by Kevin Kelly, founder of Wired magazine, will be able to multiply, by a few orders of magnitude, our capacity for analysis, curation, knowledge creation, and revision. When integrated into most companies' internal technological platforms and solutions, these same capabilities will bring savings and efficiency gains that are difficult to estimate today. Generative AI will make our current way of working obsolete, significantly changing the cost structure and, in consequence, redefining the economy of scale in most industries.

The second is the almost unlimited customization of user and customer experience. Recent research has shown that most professionals in this area already use some of the current Gen AI tools to create or revise marketing emails, social media posts, campaign images, and product and service descriptions. However, the cost of this customization is still an obstacle to achieving ambitions of this type. Analysis shows that, in some cases, Chat GPT access can cost almost ten times more than a Google equivalent. Yet, as with cloud computing, technological evolutions, and competition will, in the medium term, make this equation much more inviting.

With rewards as inviting as these, it's easy to understand the frenzy of this topic in corporate boardrooms. But how quickly will they be able to start capturing these gains at scale? Unfortunately, the answer is not 2023. And the reason is that they, and the technology itself, are not ready (yet).
Join me in witnessing this evolution.

Paulo Camara

Chief Digital Officer, CI&T