The future is coming to sharper clarity as we get past the 24 months mark of ChatGPT / GenAI technology disruption. We now have the silhouette of what the next 36 -60 months will unfold. My focus area is for-profit businesses and hence the observations are limited to that scope.
1. LLM and its ecosystem of vectordb, agents, prompt libraries etc. will become part of tech stack of every company. It will join the ranks of Cloud service provider, Data Lake/ Warehouse, ERP/CRM/CoreBusiness app etc. Each company has to decide which LLM provider and its ecosystem they will embrace.
2. LLMs market will be an installed base play. Whoever wins the LLM installed base battle with the large enterprises – will capture the biggest value from GenAI products and services.
You may think it is easy to switch between LLMs. Our experience is to the contrary. Each LLM has its own nuance of input-output. A word here or there in the prompt yields vastly different outcomes from LLM. So enterprise will stick to the devil they know as far as possible. This is no different from what we see in other tech stack plays like ERP, Database, cloud services provider etc.
3. Clicks as an unit for transactions will end and will be replaced by natural language interfaces. Before we has click, everything needed to be typed. GUI brought in clicks and it lasted three decades. So much so that when you search in google – you are served a set of blue links and you are expected to click through them.
This click paradigm will end. Users will expect to be served the answer and have to click to find answer. ChatGPT provides answer to questions and not blue links. But – look around – all our software today are click driven. All enterprise software will be reimagined to support natural language interaction. Chat is here to stay. The user will get a lot done through the chat interface – e.g. capture sales call from chat window vs. enter it in salesforce. Co-Pilot is another UX intervention that will get refined as we go along. Multi-modality will enable interesting UX experimentation to help people get a lot done while being on the move. Ask yourself – which application can you operate while you are on your daily biking activity?
4. Business Intelligence will shift from pretty dashboard to answering the question. Goldratt used to say “Information is the answer to the question asked”. In other words, until you ask a question – everything is data. It is the question that makes data into valuable information. However, today our dashboards are designed to show data in pretty format. They do not tell you waht you need to know and when you need to know it. Have you noticed that as soon as you build a dashboard – it becomes obsolete?
Business intelligence will be about helping user find needles in haystack of data. Users will be guided through decision journeys. Pretty charts will be available – but that is not the point. Every questions will be presented as per pyramid principle – the answer to the question first with unlimited ability to dig deep into causality and predicted effects.
What trends / outcomes are you seeing?
