How will AI change our work? The short answer: I don't know. The long answer… continue reading. I'll share some thoughts.
The question of how AI will change our jobs is a timely one. I already use ChatGPT, the Bing search engine, and Notion's AI features almost every day in my work to help me write emails, get up to speed with new topics, and summarize research for me.
In this article, I'll first recap where we are right now on the AI journey. Second, I'll consider the changes AI has on our work already today. Third, I'll venture into more speculative territory and share some thoughts on how AI might support and change our jobs in the maybe not so distant future.
A brief overview of AI
Okay, so AI... It's not technically a new concept. In 1956, there was an event called the Dartmouth Summer Research Project on Artificial Intelligence. This project is often seen as the birth of AI as a field. A lot of hope was placed in the field; however, not much progress was made in later years due to a lack of computational power and missing funding. Only in recent years (2018) did we start to see the rapid development of new AI models that got bigger and better. Last year, on November 30, AI hit the mainstream with the release of ChatGPT, which suddenly made it easy to interact with an AI model (here, a so-called large language model) without the need to have extensive programming knowledge.
At the moment, we speak of Artificial Intelligence (AI) as technologies that allow machines to exhibit human-like intelligence. Narrow AI refers to systems designed for specific tasks. It refers to AI systems designed for specific tasks and defined objectives. Today, most AI applications fall into this category, addressing tasks like classification problems or automating routine operations. On the other hand, strong AI, or Artificial General Intelligence (AGI), possesses human-like reasoning capabilities, including planning and decision-making. While AGI remains a distant goal, some argue that models like ChatGPT represent a significant step in that direction.
The state of AI in our workplace today
In the present scenario, AI is already making its mark in various areas of work. Companies leverage AI to forecast sales, optimize delivery routes, build customer support chatbots, and classify emails for efficient responses. These early applications demonstrate how AI can streamline processes and augment human capabilities.
AI and Machine Learning (ML), as well as Deep Learning (the latter two being subfields of AI), have the power to process large quantities of data and detect patterns in a way that humans can't. Thus, AI algorithms have helped us make sense of large quantities of text or credit card transactions, for example, to detect hate speech or fraudulent transactions way faster than humans.
Since last year, we also have ChatGPT and a number of tools built on these large language models. GitHub Co-Pilot helps us write code, ChatGPT helps us write emails and draft blog articles (I heard from a friend). Audio Pen can summarize and structure our spoken notes.
Based on large language models like GPT-4, a number of services are built. Our FLOWIT App, for instance, automatically analyzes survey answers for you. The app saves you the time you need to look through hundreds of written answers and identifies the most important topics your employees mentioned. In addition, with the help of large language models, the FLOWIT app immediately suggests measures you can take based on the most critical topics that came up in the survey.
In summary, we already see many AI tools that can be helpful in our job. These narrow AI tools do one particular task very well. However, we currently need a human at the center of the action to orchestrate all the tasks. It is difficult to say how AI will affect our workplace in the long term because we see new changes so fast. But what might we see? What would happen if we move from narrow AI to stronger general AI?
The long-term Impact of AI on our work
Now, maybe in a few years (or just months?), we will have an AI service that can connect all the different tools. Think Siri, but instead of Siri putting "grapefruit" on your shopping list when you asked it to play "Shape of You" on Spotify, Siri will now automatically write an email to your coworker to schedule a meeting to discuss the new webpage design. Thereafter, Siri will generate a list of ideas for a new webpage and ask for your feedback. If you allow it, Siri will even code a draft of the webpage after you give them feedback on the ideas. Instead of using different AI tools, you would have one that can do everything for you.
Think of an AI assistant who can also write code for you, design a new logo, do your accounting, schedule a meeting with your coworker, and book you an appointment at the hair salon. All of a sudden, we would have AI Agents who can perform complete jobs. There might be no need to hire an accountant anymore or a data scientist because the AI agent will do it for you.
Now, before I leave you with the notion that we are all going to be out of work in a few years, I believe changes take time. Maybe this situation is comparable to the invention of the computer. The invention revolutionized the job market by altering the nature of work, replacing certain jobs, but also creating new job opportunities and increasing the demand for digital skills. The computer shaped the workforce and required individuals to adapt and acquire computer-related knowledge. I believe we might be in a similar situation with AI. Already there is the new task of prompt engineering, the art of asking AI what to do so that it gives you exactly the required result.
Nonetheless, things are going to change. And from what we've seen so far, they will change fast. So, what can we do to deal with the rapid changes in the AI landscape?
- Be interested. Read and figure out how you can make AI work for you. Look up different tools, try them, have fun.
- Think about improving in skills that are complementary to AI, such as complex problem-solving, critical thinking, and fact-checking.
- Be ready for things to change and keep changing.
Disclaimer: This blog combines the thoughts and ideas of Stephan, our CTO, and Julian. We used AI to help us formulate certain paragraphs, improve the grammar, and spellcheck after completing the article.
