If one specific tool / approach seems unreasonably effective, it's usually because the problems you've been trying to solve all belong to the same category. You may not expect the same effectiveness on new kinds problems.
Show this thread
-
If all you've been doing so far is repairing rooftops and unclogging roof gutters, you may think ladders are all you need. Until you face a plumbing problem. Deep learning (differentiable parametric functions trained with gradient descent on large datasets) is exactly like that.
Show this thread
The people trying to fix plumbing issues using ladders are slightly less visionary than they think.
-
-
Just had my bathroom flooded. Used all I could. Don't you talk about ladders. What would you know?
-
It's good it wasn't the toilet. At least I exercised a bit.
- Show replies
New conversation -
-
-
People need to stand back and recognise why they are using a ladder for that job, if they are ever to develop an appreciation for the distinct requirements of plumbing tools. Most don't
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Perhaps we are more limited because we force the world to be tabular when it is more of a graph
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Also, managing to achieve "competitive performance" while using ladders is rarely relevant. It only shows how far people can go when married to an idea/tech. AI now is Ian Malcom's quote again and again (you could, but should you?).
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.