Artificial intelligence (AI) is today the most talked about technology on the planet. Having emerged from several hype cycles unscathed and with the recent explosion in the use of generative AI, the technology is being invested in and adopted at a scale not previously seen before. Its proliferation or democratization in end user segments outside the enterprise or technology world signifies that AI is here to stay and will disrupt both business and society on groundbreaking levels.
To find out more about the meteoric rise of AI, and the risks and opportunities that have emerged, HCLTech Trends & Insights spoke to David Jungwirth, Senior Director, Digital Transformation and resident AI expert at HCLTech.
AI has been around in some form for years, why has it exploded recently? What's changed and why is everyone so excited?
Artificial Intelligence (AI) has an aura of being ‘magic’. It can understand us, it can think, interact and intelligently respond in a variety of ways. All these elements have been around for decades. I remember being a very impressed adolescent, when 1998 Microsoft XP launched its first TTS (Text-to-speech) engine called ‘Sam’. Since 2010, the tech giants have battled for the best Voice-Input systems and we are all now familiar with using the next iterations of Apple’s Siri, Google’s Assistant and Amazon’s Alexa assistant for interacting with our devices, writing messages, managing our calendars, getting news pages read-out loud or receiving precise answers for precise questions, like our favorite Oscar actor's age.
Today, with the rise especially of strong Generative AIs, like ChatGPT, a huge amount of additional application areas have arisen. Receiving much longer and consistent texts, written articles or summaries, funny poems, individualized song lyrics or even professional datasheets is possible with the click of a button (see examples below). Also, the creation of images, songs or AI generated video clips is possible. A simple query in natural language is sufficient. This is a real-gamechanger as the quality of those outcomes is astonishingly good and scientific research even showed that the majority of people can’t distinguish between human and AI generated content anymore.
Is it another case of technology hype or is this the real deal?
The new generation of AI allows a kind of democratization and the broad usage of AI for everybody. ChatGPT and the underlying GPT models are more or less a framework for new specific services to be built on top of them. This has created a tremendous number of new services and applications. Founded at the end of 2022, futurepedia.io became the most comprehensive list of AI tools, reporting 20+ brand-new AI tools being launched every day, resulting in an impressive list of 2,000 AI tools across 50 categories in the market.
Is it just hype? No, I am sure AI has come to stay. We have not seen such a tremendous growth of customer facing services for many years, not even with the emergence of blockchain.
How does that shake-up today’s creative industries?
Everything in a standardized format—like a corporate datasheet—can be widely and automatically created with the push of a button. Many hours of manual, often boring work suddenly is not necessary anymore. To create outstanding and creative content with AI, a completely new profession is rapidly emerging; “Prompt Engineers” who optimize the input for such AIs. The better the input, the better the output becomes. Experienced creatives need to oversee the outcomes and ensure quality.
There also might be a shift in perspective where the creative process of AI-assisted art-creation becomes the focus of appreciation, rather than today where it mainly values the final artwork itself.
What other industries will be disrupted in the near future?
Every industry will be impacted by AI. In a recent article, I wrote about AI being a driver of digitalization across industries. One theme of this article, Customer facing AI-apps, are just the tip of the iceberg. The underlying operating model of every organization can get a massive optimization-push through AI, from operational value streams to automation and intelligent connections for software development value streams. The third theme contains the most obvious and primary uses of machine learning—intelligent data analysis, prediction and summarization—which are often left untapped by organizations. An experienced data scientist with machine learning skills can gain much better and directly transferrable insights into areas, such as the flow of value through the organization, customer habits and cost effectiveness savings, than any traditional data warehouse reporting could even deliver in its best-case.
How can AI transform society for the better?
For society as a whole, Daniela Haluza, Professor of Public Health at the Medical University of Vienna, Austria and I published four peer-reviewed AI research articles earlier this year on this subject. All today’s mega-trends are impacted by AI and in one we explored AI’s impact on healthcare. We established several AI application ideas in the healthcare sector that could be delivered through data insights, personalization for compliance, streamlining for healthcare cost savings and intelligent decision support systems.
Our latest research also revealed the strong impact of AI onto the United Nations Sustainable Development Goals.
These are two of several impressive use cases where AI can impact mankind for the better.
Furthermore, AI can help individuals, but also organizations or even states to sketch, shape and plan a desirable and equitable future.
Alright, but how does this apply to your daily work in regard to enabling today’s organizations for tomorrow's world?
Tomorrow’s organizations are required to sense market needs immediately and respond with their business and IT teams immediately. Providing customer facing releases in days rather than years requires a full change in how things get done today. Shift-Left, Continuous Testing, Continuous Delivery, (Scaled) Agile and other new ways of working are required. Today’s organizations lack the supportive tooling for bridging the “old” ways of working from yesterday, with the “new” agile ways of working of tomorrow. There is a tremendous amount of “old tooling” as well as “new tooling”, but to maintain Business Observability, intense knowledge and our capable solutions for “old” and “new” are crucial, and highly demanded in the market.
My team from Enterprise Studio at HCLTech supports this transition with organizational and process streamlining; with business observability amongst intelligent hybrid waterfall and agile planning tools and methodologies, with intelligent AIOps operations and monitoring software to eliminate alert clutter, allow automated remediation and zero-touch full-stack assurance and with AI Automation to plan, predict and accelerate Automated Workloads across distributed teams, infrastructures, network and scheduling tools.
For all the good generative AI can do, there are always risks. What are the main risks of such a disruptive technology?
The fear of job losses through AI automation is more present than ever. But the fact is different skill sets will be required in the future. For the last decade I introduced a variety of intelligent automation solutions to organizations, and typically the working people were not made redundant, instead their newly gained time was most often leveraged for much more business relevant projects. The industrial revolution as well as the digital revolution created more jobs than they eliminated.
The second argument is that AI can only be as good as its training was. Biased or discriminating data, fake-news and hate-speech are just some examples. If not properly managed, this might even lead to more discrimination than in the real world.
Thirdly, data protection and privacy are as usual of great concern. By simply chatting with an intelligent AI, we provide a huge amount of data and insights about us, just by raising questions. Earlier this year, Italy banned ChatGPT due to potential data privacy violations.
Finally, democratized AI availability for everybody and the awesome breadth of the ecosystem are just one part of the truth.
What is the other part of the awesome democratized AI-access truth?
On the one hand, publicly available, large AI models like GPT from OpenAI, drove a great democratization as mentioned before. On the other hand, they make us even more dependent on the tech giants.
Microsoft invested $14 billion into OpenAI to leverage the GPT software as an interface for their Bing search engine, as well as for interacting and automating their Office toolset. For users this is highly desirable, as it does not only output some texts, but it also really helps to do stuff and automate their tech stack based on natural language interaction. It means, not only your previous Bing search results, but also your conversations with the AI can be used to highly accurately profile your personality—fully automated by a single company.
Why did 1,000 senior tech leaders and 25,000 supporters sign a letter asking for AI development to be halted?
Large AI models are very expensive to train, and the usual tech giants are in a race to create the best AIs. Nobody, not even their creators, can fully control the outcomes and ensure that they are used for mankind’s good.
To be accurate, this famous letter does not suggest pausing AI development in general, it only suggests pausing training models larger models than the current GPT-4 AI, which is the underlying technology below ChatGPT, to avoid even more unpredictable black-box models with ever emergent capabilities.
Socio-economic norms, ethical guidelines and more predictable, positive outcomes are not developed at the same speed as new AI technologies arise. These leaders asked to pause large model trainings, until those guidelines are discussed, aligned and agreed on for some kind of AI governance framework.
Read more insights from David here