Video: New Tesla Roadster Reveal

[Runwildboys]

In 50 years, new drivers will need to get a pilot’s license

1 in 1,000,000 might pass the test. They're gonna go through a LOT of driving instructors.

 

[VaqueroTD]

Teslas are bad arse. I even love the ugly trucks.

 

[gtb1943]

In theory yes, but it doesn’t seem to be a likely scenario.

I’m not a big tech guy so I could be wrong but from my understanding it’s extremely tough to hack into these EVs and when people do it’s to do nickel and dime things like adjust the seats and window settings. Probably creates more of a theft concern than an accident concern.

did into this a little deeper
the way vehicles are controlled by their internal computers
you get access to one system you get it to all
ABS systems; the entire braking system even. Its all computer controlled.

 

[CyberB0b]

did into this a little deeper
the way vehicles are controlled by their internal computers
you get access to one system you get it to all
ABS systems; the entire braking system even. Its all computer controlled.

It’s encrypted and hard to break. Not saying it’s impossible, but the effort required would be better spent trying to hack a bank or break cryptocurrency encryption.

 

[Runwildboys]

It’s encrypted and hard to break. Not saying it’s impossible, but the effort required would be better spent trying to hack a bank or break cryptocurrency encryption.

Not if you're trying to cause an "accident".

 

[gtb1943]

It’s encrypted and hard to break. Not saying it’s impossible, but the effort required would be better spent trying to hack a bank or break cryptocurrency encryption.

so how many dead in mass accidents is acceptable?

 

[CalPolyTechnique]

It’s encrypted and hard to break. Not saying it’s impossible, but the effort required would be better spent trying to hack a bank or break cryptocurrency encryption.

Enter AI.

 

[YosemiteSam]

It's not even a concept car at this point. It's 100% a sheer marketing ploy in attempt to prop up Tesla's corporate value so he can merge Tesla in with xAI. xAI needs cash and SpaceX most certainly can't foot the xAI bill and expect to continue to operate SpaceX. Even with SpaceX's $100B in assets.

Spend $25M-$75M on a marketing stunt and suck in $500M from sucker investors into Tesla them merge xAI and loot Tesla for all it's worth in attempts to keep xAI's hopes of survival alive.

xAI between debt principal and finance leases compose ~$40B and they are looking to take another $40B in financing ($10B in bank loans and $30B in corporate bonds totaling $80B of debt after the next round of financing)

They are in the same position as OpenAI and Oracle. Enormous amounts of debt and no real way to cover that debt in the coming years. I believe OpenAI's current off-balance sheet debt is in the neighborhood of $665B. I cannot see any feasible way to cover that in the next 5-7 years. Competition is fierce and memory and CPUs/GPUs prices will start falling by 2028 and they will begin to flood the market due to all the new manufacturing plants that are coming on line and will be online by 2028.

You want to invest in AI? Invest in the small LLM model builders. The ones that build custom built-to-order model that are build to excel at very specific tasks. In the future, automation will have small task specific models that only know how to do one job and are trained on that job so there is no error rates because a model that only knows that task, is far less likely to hallucinate and because it's so much smaller (requires much less memory and GPU) it will be BLAZING fast.

Like Google Home and Amazon Alexa and similar home assistants. They will become much more full featured and everyone will have their own model at home running on local hardware. Not paying OpenAI, xAI, or even Anthropic.