Generative AI has quickly captured the imagination of millions, thanks to the way it uses a wide range of inputs to create new stories, code, images, music, video and more. For organizations seeking business value from this dynamic, it is the underlying process, not the eye-catching results, that raises the critical strategic consideration of all: what AI will mean as a pervasive computing platform.
What do I mean by the AI platform? In short, it is where business requirements and specifications for software converge at a higher rate than could be achieved in the workflows of traditional, more deterministic software. Using generative AI, anyone in an organization can envision a new customer experience and leverage an AI platform to jumpstart the building process. Even in its early days, this new platform has implications for teams, organizational needs, and customer experiences.
Fast results in new ways
Recently, the CEO of a gaming company told me his vision for a virtual concierge that would be the basis for persistent, personalized journeys across a variety of gaming environments. Traditionally, testing this idea would require engaging and merging the talents and work habits of a number of industry silos and business models, including the various creative, data collection and financial processes of arcade games, terminals and mobile apps. Since this kind of testing could take lots of time and millions of dollars, the company’s normal momentum and demands shelved this potentially groundbreaking idea.
Using generative artificial intelligence, within a few hours we had the beginnings of an engaging animated avatar with a realistic voice, using off-the-shelf services from Google and its partners along with someone from a completely different industry who was skilled in the new art of “prompting” or training the AI model to produce optimal results. The CEO’s leadership team could jump into the meatier aspects of building a new customer experience and accelerate the creative and business process by months. An actual prototype drives the next steps, not theoretical arguments, hopeful slides or received dogmas.
The gaming industry requires hundreds, if not thousands, of projects every year, so the prospect of radically cutting down the time and cost of creative brainstorming and prototyping is, ahem, game-changing. Add to that the increasing ability to use generative AI in software development, and the time from concept to code goes from months to days.
This revolutionary dynamic generalizes to almost every industry, and it’s happening now. Online travel companies like Priceline improve travel planning capabilities, retailers like Carrefour create complete marketing campaigns in minutes, and consulting organizations like Capgemini build hundreds of industry-specific use cases to streamline time-consuming business processes.
Collectively, this shows a larger pattern of increasingly easy and powerful human-computer interaction, managing complexity on an unprecedented scale, and greater enterprise awareness through improved data collection and management. All three are critical aspects of the AI-based technology platform.
Platforms and human-computer interaction
Technology platforms are transformative on a micro and macro scale for the ways in which they can reduce friction to existing processes and spur innovations that create entirely new industries. Good examples of major new technology platforms include the mainframe computer, the personal computer, the World Wide Web, mobile devices with apps, and public cloud computing. In all cases, they supported and extended the value propositions of a wide range of companies, while others were able to leverage the platform to do new things, such as desktop publishing or social media.
A notable feature of these platforms is their progression towards closer and easier connection between humans and computers. Starting with punch cards, we’ve seen transitions to command lines, drag-and-drop icons, and chatbots, along with new frameworks and languages that exploded software’s role in the world economy. AI-enabled computers will be performing increasingly sophisticated calculations as we speak.
New platform, new values
In all its forms, AI is powerful because it detects and exploits patterns. This makes it a tool that aids one of humanity’s greatest cognitive skills. Pattern insight is the foundation of the scientific method and servicing markets – our society’s two cornerstones of innovation. For example, pattern-spotting AI is at the heart of understanding how proteins fold, and this is how a generative AI service trains on an LLM and decides what to write next.
Whether it is humans or machines searching for patterns, and increasingly it will be both, the quality of the result depends on the quality of the data, to the point where rich, diverse and, above all, accurate data can be the single biggest driver of success. Serving this need will be big business in the growth of the AI platform.
Like its predecessors, the capabilities of the AI platform will improve to the point where both employees and customers will expect accurate and timely information, more efficient use of resources and personalization that changes depending on the context of the moment. It is thus a company that is not only of one pattern, but an intersection of several, at new levels of complexity and risk management.
To achieve this, traditional software developers will be able to build more sophisticated programs, increasing the system’s ability to connect, improving data quality, and increasing system efficiency. They will become greater participants in the creative process and find numerous niches to enrich.
Who wins and how
I am confident that this will happen because I see the new platform being built today. Developers at some of the world’s largest companies are looking at future environments where efficiencies such as coding prompts and model tuning increase the speed and power of existing implementations and architectures. New AI capabilities are rapidly spreading among businesses and consumers, accelerating virality and redefining the speed at which a process can be improved. They want to build within established rules, build best practices rather than risk the cost of stopping.
The knowledge they gain will not remain in silos as entrepreneurship rushes to fill the needs of a new platform. New teams and companies are addressing new platform needs such as specialized chips and data management. Hyperscale clouds provision backend systems and create new ways to meet the demand for AI processing, which can increase 10- or 100-fold overnight at a single customer. New data frameworks and multicloud technologies are already unifying data from disparate sources for better analysis and action.
Established companies have historically found it difficult to adapt to new platforms. The challenge to management is inertia along with managing cash flows from existing businesses. This time, however, companies have some platform advantages, such as rich data sets, established brands and good customer understanding. They can form alliances, make acquisitions or create wholly owned subsidiaries that bypass institutional resistance.
In addition to products, the generative AI buzz has created concerns, particularly around distortion of the use of these tools. This will become even more important as the AI platform matures. Our recently published best practices in generative AI are already design features, and together with other AI companies, we are incorporating new elements in areas such as security and watermarking. We are committed to working with a wide range of interests to build sustainable standards for the common good. The opportunity to bring these powerful tools to more of humanity, safely and transparently, is too good to pass up.
Will Grannis is vice president and chief technology officer of Google Cloud. He wrote this commentary for SiliconANGLE.
Image: geralt/Pixabay
#computing #platform #SiliconANGLE