Work is changing, and while for a while technology was there as a support we’re slowly getting to the point where it’ll not just accompagne but redefine what works means.
The evolution of AI in the last year reshuffled the cards in all industries, well to be exact it did not only reshuffled the cards it re-invented the rules of the game we’re playing.
What was at first seen as the future of work is now the future of the world:
- What happens when AI can do all your tasks?
- What is the meaning of work if you don’t need to do it?
- What happens when AI can do them better than you do?
These questions were supposed to be urgent in a decade but now with the development of concepts such as MetaGPT it’s an existential and present question.
Here’s what it’s about:
‘A groundbreaking multi-agent framework that is transforming the way software development is approached.’
As software is becoming the new way of work, you can see where we’re going.
Let’s get deeper in the reflection:
Work was originally definition of manual labour.
Then it became interaction with tools.
Then we started using machines.
Then we invented computers that can run independently.
The rapid evolution does not only automate and scale what would have required manual tasks and human intervention in the creation process but also by-pass the need of humans.
The recent few months in AI focused on a single element: How to automate tasks at scale using the power of LLMs.
AI driven tasks will dominate the way we work, and collaboration between team member + AI will be a major component of the future of work.
Something MetaGPT is clearly infused by:
The vision of automated multi-agent collaboration will be a game changer in work:
You let them collaborate and give you outputs. You can set when you’re intervention is required to verify it’s coherent with what you envisioned.
This is what the human collaboration workflows may look like.
Role-based task allocation that looks way too much like real-world job function allocation:
- Do we still need experts?
- Will we still work with complete teams?
I don’t know but what MetaGPT shows is that this vision is less valid day-by-day.
Creativity might be one of the most ‘scarce’ ressource in professional environment.
While we’re already successfully reducing the need of humans to achieve redundant tasks and operations we were far from achieving this for the creative aspect of work.
But the recent advancement in LLMs and AI foundational models gave us the ability to outsource creative tasks and to AI. These AIs gives us very close outputs to what human-expert in the corresponding field could give us.
What was supposed to be exclusive to humans is no longer.
You can collaborate with the AI at scale, get outputs, and iterate on them until you get what you were looking for.
That’s some serious magic.
Here are some use-cases:
MetaGPT concept’s are here to stay and will infuse the next generation of AI-startups.
When interacting with LLMs, ‘Hallucination’ is a real concern. As the AI is trained to please humans, there’s a real chance in some occasion the output it gives you is false.
The need for human oversight and validation will continue to be required and MANDATORY as the downside if it’s not might completely destroy the essence of the use-case.
- What will work look like?
- Who’s gonna pay the salary of who?
- What will work-life balance look like?
- Also, are our societies ready for such a shift?
- What will the future of the workforce look like?
- Aren’t some people already just working to live?
- What will be the impact on cultures and religions?
- If people don’t have money, what will they consume?
- Will the growth in productivity output the destruction of jobs?
- If AI takes a dominant stance in the workforce, what will happen?
So many questions for which I don’t have the answer but that we should all keep in mind.
What we currently need is balance, no fixed opinion on how it will save or destroy the world.