For Azure ML remote compute instances kernel picker should display only the kernels and sessions on localhost:8888

[sevillal]

Current behavior
User goes through these steps before looking at the sessions and kernels in the remote server:

Expected behavior

1. User clicks the kernel picker
2. Kernel picker directly displays the kernels and kernel sessions from the Jupyter server http://localhost:8888/ . There is no step-by-step selection, (goes straight to the kernels and sessions).
   

[DonJayamanne]

@sevillal
I'm closing this isse for now, as there haven't been any conversations in this space for over 1/2 a year I'm assuming this isn't critical anymore,
If this is, we can re-vist this, no point leaving issues when they aren't important (i'm asuming its not important as there wasn't a push for this for the past few months, since this was created)

Please feel free to create a new issue when this becomes a blocker,