A few words on AI...
In the ever-evolving landscape of technology, Generative AI (Gen AI) has emerged as a transformative force, reshaping our approach to creativity, work, and societal interaction and conversation. As we transition from conventional IT to the era of Generative AI, the question between augmenting human tasks and/or replacing jobs becomes increasingly tensed. The recent news of Klarna AI agent doing the task of 700 people comes as a solid proof of the actual disruptions in place.
The journey from simple chatbots and virtual assistants to the more sophisticated realms of ChatGPT signifies a leap towards achieving near-human-level AI conversations with brands, and disruptive startups such as Perplexity reshape web search for lunch. This revolution highlights a paradigm shift from using AI (Machine learning) for process automation-only to deploying it as a co-pilot across various platforms, enhancing human capabilities at Work and Home, and fostering a radical wave of multi-modal innovation difficult to follow.
Yes, Gen AI's potential spans across task typologies, from simple, complicated, complex to chaotic processes, demonstrating its capability to handle a wide range of activities from article writing to medical diagnostics and trend analysis. This adaptability positions Gen AI as a crucial tool for developers for instance, who leverage it as a "second brain" to enhance coding and action planning, albeit with limitations on autonomous decision-making and implementation.
Generative AI acts as a parallel system of thought and creativity. We like to see it as mimicking a logical sequence of post-its, through vectorized words. Each word is influencing others, like music notes, like each individual in a small group of people. Prompting has even become a discipline in itself with millions of approaches, each tailored to specific needs and specific models, requiring iterative tests, frustrating scientists looking for perfection and total control. Gen AI is challenging the conventional IT paradigms, very black & white and deterministic. It's also challenging design itself by coming with a general purpose tool. How to limit it, constrain it ? And why and when limit it ?
Generative AI is a success and it will last. People and businesses don't need perfection. They just need help when they are blocked, on time, in their language. The world is embracing it and its integration accelerates performance across knowledge work, pushing the boundaries beyond IT to disrupt traditional processes. Sales, Marketing, DevOps, HR, all departments are impacted through the major enterprise tools. Even regulated environments are finding solutions for better compliance, fraud detections and medical diagnosis.
IT < AI ?
The question again: Will Gen AI augment human tasks, or will it replace jobs altogether?
The answer is multifaceted. While some roles, like sales management, may see AI as an augmentation tool to take better decisions, others, particularly in call center support, might face direct replacement and drive substantial cost reductions. However, Gen AI also paves the way for more qualified and better planned meetings for sales and customer success, indicating a complex interplay between job enhancement and obsolescence. Having more time to support people and collect more feedback on products and services is a huge opportunity to improve the customer journey with personalisation for instance, and collect deep knowledge from the field.
Future will raise even more questions about the responsibility, feasibility (the size!), and viability (the price!) of advanced and closed models like GPT-5, Gemini 1.5, Mistral Large 2, and the potential realization of Human-level AI, planned in 4-5 years according OpenAI or Meta. Maybe we will stay for a while on robust and existing models GPT-4, Gemini 1, Anthropic 3, fine-tuned/RAG with your data. Maybe it will too good and "something like fire".
Meanwhile Generative AI -as we know it today- stands at the forefront of technological advancement, revolutionize how we work, create, and interact within our society. Its integration presents a true opportunity for augmenting human capabilities and creativity, improving your customer journey through knowledge and answers for instance, and also a myriad of possibilities of new value propositions that we can't wait to share with you.
The design + data + trust dialogue between your various stakeholders (technology, strategy and operations) will undoubtedly help you shape the future of your AI-powered digital and physical worlds. We are here to co-design it and keep the pace. Much more to come here.
Raphaël Briner, March 8th 2024
