Quote:
Originally Posted by Mihkel05
Oh. You just plainly know nothing about the subject matter and a poseur of the highest order. The idea that increased pixels would make any difference in feature extraction is painfully ******ed. It is obvious you made everything up to anyone with brief exposure to the field.
Have I found the TSLA fanboy?
It's clear that you have little clue on this subject. Even people in tech are usually idiots - look at the predictions of AI researchers from the 50s onwards. The fact is that pixels and the bandwidth to process them are everything for object detection. I've written moving image detection software in another field, so I'm very aware of the problems. There's an irrational optimism among idiot researchers (see the other thread) about machine learning, but the reality is a lot more sober. See this
MIT tech review article for a good overview of why the bottleneck is pixels. In controlled conditions, current resolution works just fine. Outside of it, they're inadequate. For autonomous cars to work you need much more advanced object detection than we have today, and that is impossible without much higher resolution and bandwidth than is being used today. Have a read of that article (away from Google spin) and you might begin to grasp some of the issues.
The idea that having 100+ points for reliable object recognition is not much better than several fuzzy ones is utterly ******ed. It's the edge cases that matter for commercial application, not GPS-mapped, route-planned driving on the highest quality properly signed and marked streets with a backup human. From the article (the bolded are just a small fraction of the issues that need higher resolution plus more sensors and higher bandwidth to solve):
Quote:
Among other unsolved problems, Google has yet to drive in snow, and Urmson says safety concerns preclude testing during heavy rains. Nor has it tackled big, open parking lots or multilevel garages. The car’s video cameras detect the color of a traffic light; Urmson said his team is still working to prevent them from being blinded when the sun is directly behind a light. Despite progress handling road crews, “I could construct a construction zone that could befuddle the car,” Urmson says.
Pedestrians are detected simply as moving, column-shaped blurs of pixels—meaning, Urmson agrees, that the car wouldn’t be able to spot a police officer at the side of the road frantically waving for traffic to stop.
The car’s sensors can’t tell if a road obstacle is a rock or a crumpled piece of paper, so the car will try to drive around either. Urmson also says the car can’t detect potholes or spot an uncovered manhole if it isn’t coned off.
But researchers say the unsolved problems will become increasingly difficult. For example, John Leonard, an MIT expert on autonomous driving, says he wonders about scenarios that may be beyond the capabilities of current sensors, such as making a left turn into a high-speed stream of oncoming traffic.
It's the edge cases that matter, and the edge cases depend on a myriad of secondary context cues. Today's sensor bandwidth, image quality and resolutions aren't up to deciding between these edge cases. There are plenty of idiots in IT who oversimplify problems and claim that it's just software. That's nonsense. Some of the smartest people thought AI and self driving cars would be possible soon, without realizing we were 50+ years away from the processing power and image resolution.
Now the same idiots 40 years later think that feeding low resolution data into fuzzy object typing training algorithms is going to solve the driving problem. It's not. Its going to solve a small controlled subset of the driving problem, while ignoring the edge cases that are the real obstacle to autonomous cars.
Quote:
Your question is a total strawman as well. Obviously talent poaching is an issue in any business, especially one with a well funded entrant such as Apple.
Thank you for agreeing with me. You also agree it is going to ramp up. This is certainly going to hurt TSLA's engineering efforts and very optimistic launch dates.
My contention is that increasing engineer poaching (and not just from Apple - there are a number of well funded car companies with ample spare cash) is going to cause production delays for TSLA, enough that they'll be overtaken by more capable companies with far more cash to burn, who are now entering this market. It's a classic jump ship scenario, and Apple's move is now enough pressure to guarantee it will happen.
TSLA has an enormous cash burn, less than a year of reserves, and no funding source except for selling its own stock, which requires a high stock price to not further enter a death and confidence spiral.
Quote:
However, this is is almost completely irrelevant to your thesis that Tesla can't retain engineers which will then spell the death knell for Tesla, which has clearly already pivoted into a battery company away from a car manufacturer.
So they're building a factory - for which they haven't even got the funding yet and will need to sell more stock - and you're claiming they're now a battery manufacturer? That's just laughable, dude. They haven't "clearly" done anything except build a building shell in the Nevada desert. They're making as yet unfunded bets on as yet unproven interest for as yet unproven profitability.
I'm sure you bought Musk's hype about "preorders" (Musk is a great at dishonest hyping, that idiots tend to love), but the reality is that
they're total bull**** and hype
Quote:
Your posts on tech, in general, are pure nonsense. You're the gold bug of tech companies continually posting irrational and silly thoughts about subjects you know next to nothing about.
I've made a small fortune trading tech and being correct about it often enough to make money. Have a look at my posts in this thread, just for starters. In the very first page I was arguing strongly against shorting TSLA when others thought it was hugely overvalued in the 100s; at another I was quite sure they'd beat based on VIN numbers (and they did). You on the other hand have done what? Spat vitriol?
Give me a thesis and prediction for TSLA. We'll see who's right.