I’ve noticed that leaders keep pushing for AI when we actually need more automation. AI is not deterministic and is a struggle to implement. If they want correct answers 100% of the time, why not implement automations instead? Makes no sense…
6 replies (most recent on top)
@af Yet…
It doesn't matter. With token pricing increasing dramatically, AI solutions are going to be too expensive. We'll end up using real AI - another Indian.
They want the easy magical solution that uses the cool new flashy thing.
The funny thing? AI can be leveraged by existing or new deterministic automation solutions where applicable. But the leaders making decisions aren't engineers, and have a view that applications can't evolve or add new features. That they have to build something NEW instead to show off and gain notoriety.
And so you get a cycle of projects gaining traction and getting ki-led off.
The competitive nature in which teams have to compete for funding has been a drastic mistake. It is profoundly stupid. I get it, you want teams to have some drive to make better solutions. But it doesn't work out in a company like Optum where non-technical people are even so close to engineering that they're directly managing the engineers. Instead what you get are good products and teams getting ki-led off in favor of unproven moonshot projects that exist solely to extract money from the business with no vision for the future.
I love AI - it never talks behind my back.
because "AI" is the new buzzword fad. Automation is "old"
Because you don’t get promoted for saying “hey there is a better way to do this that isn’t AI”. They will say you are not thinking 100x and that if Google can do it, we should be able to as well.
The truth is, most automation products have been being ki-led off even though they worked and continued to be improved. The business leaders want an AI product that comes in and just does the work regardless of what the work is? For example, instead of taking time to clean up data, put spreadsheets into DBs and make structured processes, they just want an AI to do all of it for a cheap price and be 100% accurate.
Leaders are incentivized to make new products instead of investing the products they already have. The leader doesn’t get a promotion for continuing to build upon a product, they need a new shiny thing they can put THEIR name on. 95% of the time the new product is worse than the old but they don’t give a sh-t because they won’t have to use it. They only get the accolades for “championing” an idea that when actually implemented is terrible.
Welcome to Optum where good ideas are ki-led for shiny BS.
This has happened over the years before AI but recently Optum has pretty much cut anyone with a brain that says “hey why would we build this thing if we already have this other thing. Why can’t we just improve what we have rather than pour millions into an unproven idea that might not work at all?”