@DanBatten@infinitegard3n Yes ! This is definitely the good approach. With cost to launch a business coming close to zero, what the market decides is even more important. Serendipity is the future for companies.
Super intéressant, le fait que ça reparte aux US renforce l’idée que nous sommes dans le creux d’une courbe en J classique (Brynjolfsson). Le seul point négatif est effectivement pour les jeunes jeunes diplômés. Lorsque l’entrée sur le marché est difficile, cela impacte souvent la carrière sur le long-terme.
@BoringBiz_ We just don’t know what the future jobs look like. Finance jobs won’t die. The transformation of businesses with AI is economics and costs analysis before everything else. We don’t know the tools and the job titles yet, but the competence is in finance teams.
AI is not killing finance and the CFO. Economics is the primitive to every business transformation. The agentic one is no different.
The first step of going to agentic is knowing befeorehand what to transform, what to prioritize, just like any project.
For priorization, controllers already have a view of what's profitable and what's not. Moreover, as there will be a lot of pushback, transforming is easier to justify on a non-profitable product or activity. That's where companies know where to start first.
A clear map of workflows is also needed, which will be the task of an FDE (forward deployed engineer) or an in-house from each department. But companies also need an map of economics next to the operational one. Each company embarking on such transformation needs to evaluate some kind of business plan of forecasted economics on the new workflows. Trade-offs and arbitration will be everywhere, and decisions will need to be approved. If the tooling for this task is not yet created, the competence surely lives in the finance team.
The prerequisite to an intelligent internal revolution is economics and controlling. As we need more engineering, we'll definitely need as much finance. Exciting time ahead for finance.
@cory_fitz@incredutility Social bonds is like one of if not the greatest variable linked to longevity. A big study was made about that so this explains a lot I guess.
We naturally tend to imagine the future of AI in businesses as an internal revolution. But we often forget that technology is driving companies to outsource functions. On SaaS and Ai workflows, do these trends just mean outsourcing?
First, regarding SaaS, the natural trend is shifting from selling a tool managed in-house (seat-based pricing) toward selling end results (outcome-based pricing). If a provider starts selling an end result, that's plain and simple outsourcing.
Second, if FDEs (forwards deployed engineers) aim to replace human processes with agent-based workflows, we might ask why each company’s customized workflows could not also be handled completely by an external provider. When it comes to handling outliers and exceptions, does this necessarily require an in-house employee? Companies will always choose the simple and complete service. Specialized providers in each industry sector will soon be fully capable to provide a 100% outsourced service to meet this need.
Let’s also not forget that the skills required for the maintenance and continuous evolution of these workflows are not found within the company but with the FDEs. And if we ever reach a point where evolutionary maintenance is managed primarily by AI, we end up with a model similar to SaaS which is delivering an outcome. Again, it's complete outsourcing.
To take a historical example from 1989, Kodak entered into an agreement with IBM to outsource its data management. It was a first at a time when companies were expected to handle this function in-house.
Changing companies from within is the logical way of seeing things. But as we're seeing AI remove friction everywhere, this could very well lead companies toward a fragmented and hyper-outsourced model, where entire 'pre-AI' divisions could become the services of tomorrow.
By trying to be correct all the time, you end up never saying anything at all.
What kept me from posting on X for months if not years is that i've always been more afraid of being noticed and come across as wrong than of remaining invisible. I've always found it way scarier to write something not perfectly right and being ridiculed by users in public when i would finally have gathered the courage to publish a text. So I ended up trying to find the perfect first post, that cannot be challenged in any way.
In the end months go by and nothing is created.
Yet being ignored is the real hurdle that awaits anyone with any kind of project. The real barrier to overcome is noise and general ignorance. It kills more projects than critics. People inherently do not care about you. That's what I should be afraid of. There is the real challenge. Whether it's launching a startup, any project, or posting on X, what matters in the end is trying to have any kind of impact on anyone.
Be seen, polarize, move matter.
@krispuckett yes. These posts need to define what "making progress" means. Cutting relationships and social life by locking in is not what I consider progress for any sane individual.
Just as AI will become increasingly prevalent in society, I’m convinced we’ll witness a renaissance for art, handcraft, and traditional ways of getting knowledge. Generating anything without effort with AI means rediscovering the value of things made by humans, and human to human relations.
Interesting use case for agents could be autonomous coordination.
Searching, analyzing and negotiating in purchasing is time consuming and will definitely be handled by agents soon enough.
But agent coordination could unlock a new dimension: purchasing power.
When thousands of agents from many different buyers coordinate together to negotiate as one buyer, things become pretty interesting. This could definitely become something big for businesses, as they always look for cheaper supplies.
Big opportunities for some kind of startup purchasing cooperative, or directly embedded in fintech or supply chain products, as they already have the users.
@housecor Agree, friction is being removed everywhere. It’s being removed in consumer business models (subscriptions, insurance) with tools like Muse. There is no reason it will survive in organizations. Human to human is where there is the most friction.
AI just revealed some established business models are bound to disappear as consumers will soon be able to easily challenge any subscription from any provider. This is what we call removing friction.
But what if AI removing friction could also profoundly impact human organizations?
Corporate bullshit jobs emerged to handle friction between people or moving information to a different level in an organization. Meetings, reports and slide decks are tools needed for human orgs. Gathering information in mails and reporting information to an upper level is not a job anymore. It’s just part of any agent’s workflow. What’s keeping all of this alive is time for orgs to change or new ones to take over.
Humans relations are everything but deterministic and we created entire levels in businesses to tackle this. However, this is hardly needed anymore. Optimizing organizations means finding what every employee is best at, while limiting or removing as much frictions as possible with its counterparts as AI is much more suited and efficient for such tasks.
We optimized organizations for centuries, made this an science we’re fond of. It has always come at the cost of what’s not efficient. This time is no different, except we’re the ones being removed.