well, the argument is, i think, in one situation the error resides on the individual, on the other the error resides in the organization. i don't think this would be really comparable to AI drivers unless each car differed significantly enough to establish them as actually separate. if i use the same equation 100 times, and its wrong every 1 time, it cant be called an accurate equation.
i think the main risk is mob mentality, especially polarizing stories could result in essentially innocent people having their private lives laid bare over sensationalist stories. you could, i suppose, couple the idea with a republican idea of senate to dampen that effect, but i suspect it wouldnt work.
seems a bit like they are punishing the loyal. id think that you could drop commercial percentage in an effort to lure viewers back and make up for the loss of revenue by time with an increase in viewership? or perhaps the commercial boat has already sailed and a new revenue model is needed.
Yeah, the presentation format is annoying but argument is good, in fact, the argument is really important.
Machine learning is about approximate reasoning but with no real guarantees of the approximation. Even if the approximation is usually very good, being occasionally bad can be "deeply problematic" when we don't have control over exactly how exactly the bad recommendations are created.
I'd say that's more what cost functions are for, but that is neither here nor there.
I think the biggest takeaways are...
1) Machine learning in large software projects is complex because decisions made by one algorithm can influence the data of other algorithms, creating massive biasing.
2) Simple cross-validation / hold back methodology is limited as we expand what we want machine learning to be able to handle. Reason: big data is too big and "correctness" is difficult to evaluate for things like Q/A systems.
Indeed it is and machine learning not being "real statistics" but ad-hoc methodologies (based on statistics but not being statistics) doesn't give confidence intervals etc.
Plus read the article/presentation. It goes into the issue with great depth and clarity.
If we pick a silly thing like recommendations, they happen in a multidimensional space. Not only do we not know the right metrics, shape of the space or cost functions, our feedback is distorted by the actual recommendations we give!
Doesn't matter. Elon still relies on government funding because he's selling to government.
It's no different than when Rockwell/North American built the Space Shuttle.
The biggest problem is that his ideas are just so boring and uninspiring. How about a new idea for a change? (Hyperloop isn't more exciting than High-Speed rail)
Being known for someone that "did something for slightly cheaper" isn't exactly awe-inspiring. Sucks that millennials are stuck with him for inspiration.
SpaceX's manifest for this year [1] shows 7 US government launches, 6 commercial launches, and 1 SpaceX-paid R&D launch (Falcon Heavy). Yes the US government is pre-paying for some R&D for Crew Dragon, but no, it doesn't add up to 90% of revenue.
>Everything they're ... planning on doing, has already been done.
So landing a rocket after launch to re-use that rocket has already been done? Going to be interesting when they are the first to do so but it's already been done. I'd love to hear how you reason that.
Doing things in a more efficient, cheaper manner means innovating new ways of development, testing, and building. More 'affordable' ventures into space can lead to an array of improvements. More research satellites being sent into space, as an example, because instead of costing $12 million to ship into space it costs $4 million [0].
I'm sorry you don't see the value in such things, but rest assured it is there.
[0] figures made up, can't be bothered searching for accurate pricing
What other nation or private entity has come up with a visionary idea in space industry? Let me remind you he first thought about sending a life capsule to Mars on a former ballistic missile. Isn't that visionary enough ya?
