Generative AI has burst onto the scene as a hot topic in the public debate, capturing everyone’s attention. However, it is worth remembering that the technological bases of this revolution are not new. The Internet, smartphones, and powerful computer chips have been evolving for decades to pave the way for this exciting future.
Until now, the first manifestations of AI have mainly focused on tasks like virtual assistants, machine translation, and search engine enhancements. But now, with the growing power of generative AI, we find ourselves at a fascinating crossroads.
Its tools promise to boost human learning and productivity, applying big data in probably every industry imaginable. In addition, the threshold for creativity is being lowered, putting the power of new computer code in everyone’s hands.
“This massive use of generative AI not only promises to transform computer vision and natural language understanding, but also has the potential to accelerate the development of technologies that enable autonomous vehicles and robots to make automated decisions,” exone in a report Marco Barresi, equity analyst at Lombard Odier. “We are facing a true revolution that redefines the limits of the possible”, he adds.
The expert assures that despite the promises and threats surrounding AI, companies are still in the experimental stage of its adoption. “This creates a challenging landscape for investors looking to identify successful business models and assess the profitability of this technology,” he says. In his view, the fervor and interest in AI is unprecedented, which is sure to fuel the emergence of new solutions and accelerate the pace of adoption in the near future.
Take the example of Nvidia, the American computer chip giant, whose sales and profit projections through 2024 surprised the markets. Barresi says their highly sophisticated chips, critical to powering AI models in data centers, have catapulted them to a market capitalization of close to $1 trillion. “This historic achievement positions Nvidia just behind Amazon and ahead of Meta (formerly known as Facebook),” he says.
Although the broader semiconductor market has faced challenges, such as high inventories and slower demand for less sophisticated electronics, the outlook for GPUs, Nvidia’s specialty, is entirely different, from his point of view.
Barresi says that these highly specialized graphics processing units are vital to the development of generative AI, driving advances in semiconductors for computation, storage and transmission of data.
“We are at a critical moment, in which generative AI is beginning to unfold its full potential,” he says. “It is time to keep an eye on the advances and opportunities that will arise in this new technological paradigm, which will undoubtedly redefine the way we live, work and create”, he analyzed.
walking to the rule
In the fascinating world of generative AI, once machine learning models are perturbed, much of the data processing can be delegated to more efficient and energy-efficient microprocessors.
However, this efficiency does not prevent the fact that data centers continue to consume an enormous amount of energy, approximately 1% of the world’s energy. The massive demand for power is driving cloud providers to constantly improve the hardware that runs these data centers.
Nvidia is an example of the underlying infrastructure that makes AI applications possible and has led the first wave of value creation in this technology. But now we are on the cusp of a second wave, where generative AI is offered as a service bundled with software products and in enhanced versions of existing applications. Vendors will charge a premium for these cutting-edge solutions that will run on PCs and smartphones.
As we move towards this new horizon, more regulation is inevitable. AI poses challenges that go beyond job substitution and “creative destruction.” There are concerns around confidence, accuracy, intellectual property protection, and data bias.
“Future legislation will need to address these inherent weaknesses of AI, especially in critical areas like autonomous vehicles and medical applications, where accountability and transparency are critical,” Barresi stresses.
According to Evli, a more formidable threat comes from the fact that “everything the AI produces is based on instructions applied by its programmers.” For example, would an AI programmed by Russia or China be credible? At the dawn of the internet age, there was a time when everyone thought the internet would spread democracy and limit state control over individuals. “Instead, the reality is that the internet has perhaps fostered division, because algorithms produce the kind of news that they think readers might like,” he points out.
The regulation will also need to address “AI alignment”, that is, making sure that generative AI conforms to human values and avoids harmful dangerous responses. “It is undeniable that these powerful tools can also be used as weapons or in disinformation campaigns and cyber attacks,” says the expert.
The G7 summit recently took a first step by agreeing to create a periodic working group known as the “Hiroshima AI Process”, in recognition of the venue where the meeting was held.
As generative AI moves towards its full potential, experts believe we need to stabilize vigilantes and ensure that the technology is developed in a responsible and ethical manner. “Proper regulation will be crucial to address the challenges and ensure that AI is a driving force for human well-being and the progress of our society,” Barresi concludes.