Generative AI
In the annals of the history of science, names such as Isaac Newton stand out, whose groundbreaking theories changed the foundations of physics. Behind these extraordinary geniuses, there were often lesser-known individuals with amazing abilities, the so-called savants. One such example is Rüdiger Gamm, a German savant and a math genius. Gamm can solve complex math problems that would challenge even the most powerful computers in a matter of seconds and keep intermediate results in mind from one day to the next. But what if every organization had access to such “superhuman” capabilities?
Imagine an assistant who carries all of a company's information in her “brain.” She remembers the 1998 balance sheet and the underlying data, the marketing strategy for 2030 and the content of today's newspapers and all newspapers ever published. She follows everything that is posted on social media channels. And not only that — it can also use all this information for complex analyses. The answer to this futuristic vision could lie in the world of Large Language Models (LLMs).
LLMs are a form of artificial intelligence that is able to analyse, understand and generate massive amounts of text data. They can perform analyses in seconds that would normally cost specialized teams days, weeks, or even months. The conclusion that you save time and money is easy to draw.
In a rather bold forecast, Goldman Sachs hypothesizes that generative AI could potentially increase global GDP by a remarkable 7% — a remarkable effect for a single technology. But these potentials are not just science fiction or visions of the future. With today's Large Language Models (LLMs), much of what sounds like a fantasy is already possible. And the models of the future will be able to carry out even more complex analyses.
As in a wave of technological advancement, visionary market leaders such as Telefónica, Mercedes and Bloomberg are already relying on the tremendous power of generative artificial intelligence and large language models (LLMs). This early departure into a new burgeoning era of innovation is driven by four central motives which, like invisible force fields, determine the strategic direction of travel.
In the vibrant world of digital technologies, the innovation cycle is like a rapid time-lapse. Anyone who misses the connection here quickly loses themselves in the fog of insignificance. With a bold leap to the forefront of technological development, companies are securing decisive competitive advantages and a position in the pole position of digital transformation through the early implementation of generative AIs and LLMs.
As if in an intensive game of chess for the brightest minds of our time, the hunt for the best talents has begun. This competition is reflected in the rapidly growing number of job advertisements for AI experts. Companies that recognize and use the potential of generative AI and LLMs present themselves as attractive employers and attract the brightest stars in the talent sky.
Generative AI is a subversive, a landshaker that has the potential to turn entire industries upside down. Even tech giants like Google are confronted with the disruptive power of this technology, as the up-and-coming company OpenAI impressively demonstrates. As in a giant chess game, the moves are redefined and the balance of power is re-explored.
In a world where automation is becoming the norm, customer expectations are also shifting. Not only do they expect lower prices, but also an improved customer experience. Virtual assistants, chatbots, and similar technologies, driven by LLMs, are setting new standards in customer interaction. Companies that do not meet these changing expectations risk losing contact with their customers.
The race for supremacy in generative AI has been launched. It remains to be seen who will ultimately come out on top and how the technological terrain will continue to shape. But one thing is certain: Those who boldly move forward now and actively set their course for the future have the best chance of achieving the goal as winners.
In our opinion, many large companies also make the mistake of investing in generative AI, but this is primarily about seniors and only very little about master's students, working students and interns. Companies that want to fully utilize the incredible power of LLMs should make these technologies a strategic priority. There are several important aspects to consider:
LLMs should not only be seen as another technology in the company, but as a central part of the corporate strategy. The integration of LLMs should be closely linked to the company's goals and purpose and reflect the added value that this technology can deliver.
Implementing LLMs requires careful consideration of data protection and compliance. It is imperative that companies ensure that the use of LLMs complies with regulatory requirements and best practices, particularly when it comes to handling sensitive information.
Companies should both build up internal know-how and draw on external expertise to get the most out of LLMs. This means that employees must not only be trained to use the models effectively, but also to understand how they can be applied to various business processes and challenges.
The infrastructure and processes used to support LLMs should be flexible and scalable. They should be designed in such a way that they can be adapted to the changing needs of the company and to advancements in LLM technology.
Implementing LLMs is not only a technological challenge, but also a cultural one. Organizations must be prepared to implement change management strategies and foster a culture of acceptance and adaptability to ensure the success of this revolutionary technology.
Like any technology, LLMs can be used for both good and bad. It is important to incorporate ethical considerations into the implementation process from the outset and to ensure responsible use of LLMs. For example, it could be important to develop ethical guidelines for the use of LLMs and to ensure that these models are not used to manipulate or disseminate false information.
For the successful use of LLMs, it is crucial that end users — whether they be the company's employees or its customers — accept and use the technology. The focus should therefore be on creating a positive user experience, for example through intuitive design and usability of applications based on LLMs.
As with any technology, it is important for LLMs to continuously evaluate their performance and benefits. This includes setting appropriate metrics to measure success and establishing feedback loops to continuously improve the models and their application.
When implementing LLMs, it can be beneficial to involve experts from various fields, including data science, linguistics, ethics, law, and business administration. Such interdisciplinary collaboration can help to better understand and address the challenges and opportunities associated with LLMs.
At Perelyn, we work with them to classify our customers in a 5x5 matrix that relates the five phases of IT/AI infrastructure maturity (Initial, Managed, Defined, Quantitatively Managed, Optimising) with the five levels of Generative AI Strategy Maturity (Awareness, Understanding, Capability, Proficiency, Leadership). This matrix could serve as a roadmap showing the company's progress in terms of technical infrastructure and the ability to implement and use generative AI models. It can help determine the current state of the business and plan the next steps to improve the use of LLMs. Such a structured approach can ensure that the company continuously develops and makes full use of the enormous opportunities that LLMs offer.
The applicability of the cells in the 5x5 matrix to companies of various sizes depends on the specific circumstances of each company, including its current technological maturity and strategic orientation. In general, companies can often be located in the matrix based on their size as follows:
Small businesses are likely in the early stages of the matrix. They may have a basic IT/AI infrastructure (Levels 1-2: Initial, Managed) and are aware of the potential of generative AI models or are just beginning to understand them (Levels 1-2: Awareness, Understanding). As a result, the cells that represent this intersection are best for small businesses.
Medium-sized companies may have a more advanced IT/AI infrastructure and be able to use specific AI features (Levels 2-3: Managed, Defined). They could have a better understanding of generative AI models and build their skills in this area (levels 2—3: Understanding, Capability). The cells that represent this intersection would therefore be suitable for medium-sized companies.
Large companies, particularly those in technology-intensive industries, could be in the higher levels of the matrix. You may have a well-defined and quantitatively controlled IT/AI infrastructure (Levels 3-5: Defined, Quantitatively Managed, Optimising) and are on the path to mastering or leadership in generative AI strategies (Levels 3-5: Capability, Proficiency, Leadership). Therefore, the cells that represent this intersection would be suitable for large companies.
The journey to implementing Large Language Models (LLMs) is like a journey into a new era, a departure into unknown territory that is rich in opportunities and challenges at the same time. When we imagine savants like Rüdiger Gamm, who can calculate with impressive accuracy and speed, we see a picture of the enormous potential that can be unleashed by combining human intelligence and technological innovation.
The “superhuman assistant,” this almost mystical idea of an artificial intelligence that can learn, adapt and act in close cooperation with us, suddenly doesn't seem so utopian anymore. Every progress we make in AI research brings us closer to this goal.
Introducing LLMs can seem like stepping into a new era. But as the work of savants such as Rüdiger Gamm shows, only the combination of human potential and technological innovation can deliver breakthrough results. The “superhuman assistant” in companies could soon no longer be a fantasy, but a valuable resource for companies all over the world.
The world of generative artificial intelligence presents many challenges - we at Perelyn are your experienced companions on this exciting journey.
We have completed over 20 successful generative AI projects in various industries (e.g. logistics, media, manufacturing, pharmaceuticals) and functions (marketing, customer service, research & development), and carried out a total of more than 200 AI projects for medium to large companies.
Our adaptive pricing and ability to utilize near-shore capacity helps you manage costs efficiently.
We have a network of over 50 experts in DACH, Scandinavia and Eastern Europe, ready to make your project a success.
We are the first company in Germany to offer a comprehensive strategic approach for generative AI.
As a leading thought leader in the field of generative AI and large language models, we are recognized and supported by leading companies and organizations.
In our latest article, we highlight the central role of compliance and quality management in the development of AI systems. We show how innovative approaches can be reconciled with regulatory requirements — an important step for companies that want to drive sustainable innovations in the fast-moving world of technology.
Discover the impact of generative AI in retail. This technology unites departments, personalises content, and transforms customer experiences. Leading brands such as Coca-Cola and Walmart are already using their potential to optimise operations and drive innovation. Explore the future of retail with generative AI...
This article explores how large language models such as ChatGPT can be integrated with proprietary data. The recent advent of large language models opens up numerous opportunities for companies, but connecting with proprietary data is a challenge in such a transformation. This article discusses the various approaches to solve this problem, such as fine-tuning and contextual learning. In addition, the associated challenges and risks are addressed.
