donniemattingly

3 ms·

1. 1.▲
   by donniemattingly 4y ago

   > Second, then the safety around the model. Ad runtime. We have lots of classifiers around harmful content or bias, which we then catch. And then, of course, the takedown. Ultimately, in the application layer, you also have more of the s
2. 2.▲
   by donniemattingly 4y ago

   seems like microsoft has multiple layers of ‘safety’ built in (Satya Nadella mentioned on a decoder interview last week). My read on what’s going on is that the output is being classified by another model in realtime which is then deleted i
3. 3.▲
   by donniemattingly 4y ago

   If you’re not interested in DIY shapeoko and x-carve are options well under $5k as well. I second the MPCNC recommendation. Though if your goal is production and not tinkering with a CNC machine, it may not be the best fit.
4. 4.▲
   by donniemattingly 6y ago

   These questions often ignore the investment (both in time and money) people have made in a particular platform. Even if you assume each platform is equivalent (which isn't true), there's still switching costs incurred moving from