https://www.kickstarter.com/projects/974179222/ability1-3d-print-real-metal-parts-at-home/rewards
really cool- smart idea- it must take a long time to lay down a metal layer then machine each layer- still does produce metal 3D printed parts.
Very cool solution, wish I had the money.
Ok, so they’ve fixed the “this will literally kill you” problems. Their initial prototypes had no safety consideration for argon purge indoors (which will asphyxiate you) or desktop arc welding (which can shock or blind you). So, that’s progress.
But… WHAT’S THE POINT? It basically can’t make any part geometry that a CNC mill can’t make. (There’s no bridging with MIG welding…) Why not just use a mill?
@Ryan_Carlyle There is no point. It could be kind of fun to tinker with a machine like that though.
But it will probably be another kickstarter failure.
It’s also limited by the milling cutter diameter, so you cant fabricate parts with small holes in the middle. Supports will be a nightmare as well, presumably. All in all, nice concept, nothing new - it’s been done before with a laser instead of a MIG welder - not really all that useful for that kind of money. I think @Rien_Stouten is correct, in that this one will bite the dust.
i think the idea is understandable coming from FDM (metal wire = filament) but i don’t think is very practical/efficient in the long run, there are better methods out there using metal powders and either using binders and heating etc that produce better and faster results, it may be a “new desktop” installment but 3+k is still pretty expensive. I’d say keep working on the idea but in no way at this point in time does this look like its kickstarter ready/stable and will likely be another “failed” attempt to put on the other pile of failures from kickstarter. With that said I got my #polysher today! woot! 
The price doesn’t seem like a problem for what goes into it, but I question the viability and I wouldn’t use a welding based machine in a home or desktop setting. It should have been shown only in a garage setting.