amarka

What's wild to me is that today there are more companies with more than 100 unsupervised self driving cars out on the road that DID NOT exist (as a company) back when Tesla CEO told investors they're a year away from FSD.

True story, for example take Tesla's promise of self driving which was delivered back in 2015.

Maybe if you think really hard you’ll come up with other (more useful) examples.

“ one way that humans have historically accessed the ineffable” that’s my question, what about the other ways that have been automated away. Is there something special about mathematics or is it the authors first encounter with obsolescence?

Apologies is this is a simpleton question, but how does this differ than most of the mechanical work previously done by hand, but is now automated?

"the slowness is a feature" - I grew up in eastern block of Europe in the 80-90s, this type of sentiment has a real "the party loves you, this is for your own good" vibe.

Every copy is better than the original, true story.

Ohh sure, its users use it for all sorts of things

https://arstechnica.com/tech-policy/2026/03/elon-musks-xai-s...

Or in my context:

> Give a person code, and you help them for a day. Teach them to code, and you frustrate them for a lifetime.

While the author’s (banker and a data scientist) experience is clearly valuable, it is unclear whether it alone is sufficient to support the broader claims made. Engineering conclusions typically benefit from data beyond individual observation.

Let me fill you in on a secret: one of the reasons behind US’s dominance is that even though “there are other countries” that could do the hard science, the US is one of the only ones that did. Now the US is at risk of becoming just another one of those “other countries“.

From the article: “ The White House, which was responsible for the assessments, said the information will be housed within NASA to comply with the law, but gave no further details.”

It doesn’t seem like the admin believes the reports are biased and need to be removed. Cool conspiracy though.

Not sure if this is accurate. It seems they’re refusing to take the risk.

They are indeed some of the best floors to roll on.

I can’t believe Justin Trudeau would lowball your friend like that.

I get about 21 hours from mine, it's running Windows but powered off.

Ahh yes the "we'll rewrite this code from scratch" argument; known well to work every time!

Not sure if the last sentence is in jest or not, but here's a list of crashes for just one of their plane models:

https://en.wikipedia.org/wiki/List_of_accidents_and_incident...

The road to hell is littered with "free ice cream this way =>" signs.

Indeed it's totally worthless data, like what are you going to do, dissect people into groups that you could heavily target and try to sell them gizmos or swing an election? Pfft, not worth the effort. There are no companies or state actors are into that kind of thing.

What's worse is that I heard rumors they don't allow slave labor either. Labor laws, ugh, they're the worst.

Waymo can do inner city driving better than Tesla and GM can do highway driving better than Tesla.

I love the competition amongst the different platforms, as a consumer it's a win for us. I could care less on which brand does it better.

Why should I trust Denise Minger over Colin Campbell?

I read Capbell's China Study and I'm definitely interested in counter points to his findings, but it seems like Minger is a food blogger (and recent book author) while the other is a biochemist with a long history in their field.

My understanding is that the majority of his personal net worth is from Tesla (+Solar City) and SpaceX. Would that limit his ability to personally fund Tesla?

dsnuh - I haven't seen anyone argue against "[software+hardware] would beat a human behind the wheel", its just that developing the tech to achieve this feat with the same level of accuracy as the top 10-20% of current CDL drives is extremely expensive and caries a lot of risk. Nobody seems to of cracked the nut on it.

You're right in that there are many ways to do this, but none have come even close to beating a decent human behind the wheel.

That's a really good point. I also wonder if the cyclist accident changed the mood and/or culture at Uber when it comes to autonomous vehicles. I have to be honest that it would absolutely gut me inside to know that my software caused a death.

I think you trade one complexity (fixed vs variable routes) for many other complexities (larger vehicle with more demands). Think of the complexity of getting a driver license for a passenger sized car vs a commercial driver license, there's a lot more rules (physics and human regulations) around a CDL. So yes, you have a fixed route, but is that really the hard problem to solve, the route to take? Isn't the hard part monitoring road conditions, identifying objects, stopping for emergencies, etc?

I don't mean to be sarcastic here, but I hear developers (myself included) throw out "it would be much easier to..." or "this should only take an hour to automate..." all day long.

If it was an easy and profitable problem to solve it would have been solved by now. I agree that it would be very beneficial for everyone if trucks/truck routes were automated (except maybe the truck drivers getting laid off) but it's obviously a very hard and very risky problem to automate.

I agree that there's a case for improved efficiency, but I'm not sure if the economies of developing all the tech from scratch adds up. One accident can endanger the entire company, a problem they're still reeling from with the bicyclist fatality. The car they were trying to automate has much better stopping (and overall safety) performance than any truck, yet they're still experiencing all the woes of developing new tech. Now imagine an automated truck accidentally punting a Grayhound buss full of band kids off the road.

What do you mean by "most economical sense to automate"? It seems like everything about automating a truck would be more expensive (cost of 1 truck = 5-6 cars), the insurance and maintenance on the truck and hardware, the potential for catastrophic failure (1 runaway car doing 70 = bad, 1 runaway truck doing 70 = REALLY bad), the extra complexity of pulling a trailer and monitoring the trailer as well as the truck, monitoring the size and type of load of the truck and modifying the driving characteristics to match the load, the extra regulations that trucks are subject to (what roads they can be on, what loads they could cary on certain roads, what time of day they can drive on said roads, when they can or can't use engine braking).

Edit: even on 'simple' point A to point B route that involve 99% highway, what happens when a small part of the highway shuts down for whatever reason (flooding, multi-lan accident, fire, etc) and all traffic is routed on smaller adjacent streets, or forced to share a lane with oncoming traffic. Automating a truck is as hard or harder than automating a car, since you're dealing with the same external variables but have more internal/attached moving parts.