The way I see autonomous driving is trough reality, not trough fanboism or 'tech fetishism'.
The current stable driving assistant systems must be secondary to real driving.
They are capable of critical prevention and effective safe-nets and when done properly they reduce risk tenfold.
Idea of beta-testing on the public roads by creating consumer disinformation is outrageous and must be stopped.
It is beyond believe, how regulations are soft towards applications of technologies with direct effect on human lives. My area of expertise is in UI/UX and the simplest logic possible is called ergonomic prerequisite.
Every system must exist as a 100% functionality before some form of control surface is introduced. Every control surface must be human oriented. In car UX, the main function is driving and everything that reduces risk of removing focus from the road is good UX, everything that creates distraction and complication is not only bad UX - it is dangerous.
In car touch interfaces are exciting for people who hate driving and love their phones. This is the main selling point. It is the result of marketing driven design and enormous amount of production savings (when removing physical controls for dedicated functions) as a business motivational vector.
This will not work long-term. The Idea of vehicles as a SaaS platform also will not work. This will be CX hell and people will reject it.
> Idea of beta-testing on the public roads by creating consumer disinformation is outrageous and must be stopped.
Everyone driving the FSD Beta knows pretty well what they're doing and what the whole thing means, hence the very strict criteria the drivers must meet to have that feature enabled, so I don't know what disinformation you're talking about.
Calling it outrageous, dangerous, etc doesn't make it so, and evidently, after many month of FSD Beta, there is yet an accident to be reported resulting from use of it, in a world where every Tesla accident makes it to front page of the news. Is it an accident waiting to happen? Well, every time you get into your car, you're entering the accident lottery. It's all about the likelihood of getting into an accident and so far numbers are in Tesla's favor.
And how do you imagine a system whose improvements relies heavily on real world data is meant to obtain it without large scale participation? Have a hundred employees doing rounds in a test track?
The way I see autonomous driving through being well informed, and not speaking based on outdated intuitions, Tesla's pushing the envelope just enough to make meaningful progress toward full self-driving while not overly putting people at risk.
> It is beyond believe, how regulations are soft towards applications of technologies with direct effect on human lives.
Heavily regulating something that's in its infancy is the surest way to kill it. Those people who are constantly bitching about why robo-taxis aren't here yet even with lax regulations, can be certain to dream of it for the rest of their lives should walls be put up too early.
> In car touch interfaces are exciting for people who hate driving and love their phones.
Not only. I drive a M3 and I love driving it. I don't use my phone and nor find the urge to. I enjoy the lack of clutter, and the beautiful large screen when I need to use it. Every other car I sit in, feels like a 60s Russian rocket and stresses me out. Even Tesla's own Model S and Model X feels too cluttered now with the extra driver's dash screen.
> This will not work long-term. The Idea of vehicles as a SaaS platform also will not work. This will be CX hell and people will reject it.
Ye, that's what they said about SaaS on the Internet. Who would, in their right minds, give up control of their service and store their data and code at the mercy of a third party. Saw how that turned out.
> Everyone driving the FSD Beta knows pretty well what they're doing and what the whole thing means.
An example/analogy. If I am in the position of creating technology for automation I will implement the proven practices in aerospace industry, you have plenty of innovation and effective (with some exceptions in recent years, hello Boeing) regulatory process.
Testing autonomous vehicles on the public road is dangerous and non productive in any way or form (except gathering data cheaply).
> And how do you imagine a system whose improvements relies heavily on real world data is meant to obtain it without large scale participation? Have a hundred employees doing rounds in a test track?
Exactly. Last time I checked R&D is corporate responsibility not public data exploit or "free work". People don't get it at all.
If you will use my driving data, you must pay me or give me some incentive. I work for money, you are selling expensive product so how about half the price?
> Heavily regulating something that's in its infancy is the surest way to kill it. Those people who are constantly bitching about why robo-taxis aren't here yet even with lax regulations, can be certain to dream of it for the rest of their lives should walls be put up too early.
The technology must be produced and tested to death, before is introduced to public roads. Touch interfaces implemented without any concern of ergonomic prerequisite are proven distraction and risk.
>Ye, that's what they said about SaaS on the Internet. Who would, in their right minds, give up control of their service and store their data and code at the mercy of a third party. Saw how that turned out.
Again, the big and bold SaaS brush, yep I see how this is turning out. You can check the sales stat of NAS servers in recent years or resurrection of the flip phone.
People are stupid, until they are not. Information Security and Control will be the defining factor for businesses in the near future.
Yes, some people got madly rich on marketing Cloud and AI scams and "user generated content", so what?
This sounds as Davos idea of the future. You will own nothing and you will be happy.
Sorry, but not at all.
SaaS services have use-cases. Moving everything to SaaS model is clear madness.
> Testing autonomous vehicles on the public road is dangerous and non productive in any way or form
Where are the accidents after many months? None, so nope, not dangerous just because you think it is.
> Exactly. Last time I checked R&D is corporate responsibility not public data exploit or "free work". People don't get it at all. If you will use my driving data, you must pay me or give me some incentive. I work for money, you are selling expensive product so how about half the price?
Oh so it's money problem now. If Tesla paid people for Beta testing, it'd be fine? So you're ok for people to allegedly (which I don't agree with anyway, but most of the rest of you do) risk each other's lives as long as they get paid.
Well, you can say the same about Apple, Google and pretty much any other company that runs Beta testing. People rush to do it for free and wait in line. They are compensated by having access to latest and greatest without waiting.
You think getting paid means they are not exploited? How about paying you for you kidney. That's OK? You don't seem to understand what being exploited means.
> The technology must be produced and tested to death, before is introduced to public roads. Touch interfaces implemented without any concern of ergonomic prerequisite are proven distraction and risk.
Stop with the blanket statements. Where are you numbers that shows Tesla's are getting into more accidents? You don't have it because it's not only there but it's the opposite. Tesla's get into far less accidents that general fleet (those numbers on Tesla's report).
> You can check the sales stat of NAS servers in recent years or resurrection of the flip phone.
You look at a jump in NAS servers and flip phones and conclude that they're going to take over SaaS and slabs? Go look at the relative jump and magnitude of SaaS increases too.
> Yes, some people got madly rich on marketing Cloud and AI scams and "user generated content", so what?
Ye, the fact that a one person with a few bucks can now access what was only available to few multi-nationsal a decade ago is a scam. Do you also think the Earth is flat and COVID is a hoax?
> Moving everything to SaaS model is clear madness.
If you're actually trying to do something of value, and someone takes off a huge payload off your shoulder so you can get there faster, you'd be a fool to not do it. Time is the ultimate currency here and anything that buys you time is a win. So in principle, SaaS, IaaS and any other XaaS is a win.
The only time it's not a win is if one is either mandated to keep sensitive data in-house, or runs such a large/custom workload that no other provider can economically meet.
Man, I see you don't understand my argument at all.
You think that I care about Tesla?
Tesla is just a recent example of pushing the limits of regulation.
No. I care for human centered design.
I care about technology with responsibility without data exploitation. I care about privacy and transparency.
Your arguments are apologetic in my view and confirm the state of current affairs.
As a consumer I reserve my right to vote with my wallet.
And I will never succumb to this perverse product design philosophy that SaaS model is trying to propagate outside the effective and usable territory.
>If you're actually trying to do something of value, and someone takes off a huge payload off your shoulder so you can get there faster, you'd be a fool to not do it. Time is the ultimate currency here and anything that buys you time is a win. So in principle, SaaS, IaaS and any other XaaS is a win.
There is no "professional" lingo that can hide the core business ethics behind this race for "eating and dominating the world".
Faster? For what? May be tomorrow you will rationalize mass control with chip implants, because it is "progress" and this will get you faster toward the ultimate life goal : corporate profits.
What a dream. We will see how this "futuristic projections" unfolds.
Remember: Implementation is a key to success.:)
> I care for human centered design. I care about technology with responsibility without data exploitation. I care about privacy and transparency.
That is idealism, definitely necessary in small quantities but what moves things forward is pragmatism. The former is more of a guide than a prescription. E.g. human centered design is costly - when you're pushing the envelope, you can't have everything in utmost polish (and a mere glance at any part of history of new tech should make that obvious).
Responsibility without data exploitation: again, pragmatism - if Tesla is still doing this in 10 years, then ye, I'd find it questionable, but right now, Tesla is the only entity on which to build hope on, if you actually believe and care about what they're trying to achieve. If you don't or think it's all for Elon's pocket, then you're mistaken (with plenty of proof around) and I hope you spend some time to challenge your beliefs.
Privacy: this one's actually a balancing act. You want perfect privacy, go in the woods and hunt. In SaaS case, you trade some privacy for time and money saved.
> Your arguments are apologetic in my view.
I'm not sure how to express it without it sounding that way!
> Faster? For what? May be tomorrow you will rationalize mass control with chip implants, because it is "progress" and this will get you faster toward the ultimate life goal : corporate profits.
Pick 10 things your life depends on today (some hints: medicine and medical tech, transport, ...) and you'd be hard pressed to find any of them that don't trace back to so called "corporate profits" in capitalist nations. It's not the ultimate goal as you put it. It's the requirements to be in the game. If you wanna make large scale impact, you need capital, which comes from investors whom you need to keep happy. That's not to say no one goes in for the money, but if you spot a smart person who's earning it the hard way, it should be a good signal that they're not doing it for the money (hint: leading the first publicly traded car company in over half a century where every other ones in the country have gone bankrupt - many easier paths to making butt load of money faster).
I am not at all idealist. I am pragmatist to the core. Data is the new petrol. And exactly as in the petrol starting times, there are "entrepreneurs" and others who have no idea about value of personal data.
Moving forward everyone will have this idea, and will not share freely like in this moment. People don't care, "Because they have nothing to hide", wait and see what they will do when realize how someone is profiting over this big time. People are stupid until they are not.
Second point: Security.
All our software is big security risk. We don't have a single os created with memory safe language in mind. In the future everything will be hackable and exploitable.
You can check what is going on in Apple camp, today. Not tomorrow.
There are powers in the corporate world who don't want to see this as a problem, they care about marketing, sales and profit.
All of this will crumble to the ground.
In near future treats will be bigger, that's why your idea is "optimistic" and mine is realistic.
On human centered design....I am not interested to argue at all.
The current stable driving assistant systems must be secondary to real driving. They are capable of critical prevention and effective safe-nets and when done properly they reduce risk tenfold. Idea of beta-testing on the public roads by creating consumer disinformation is outrageous and must be stopped.
It is beyond believe, how regulations are soft towards applications of technologies with direct effect on human lives. My area of expertise is in UI/UX and the simplest logic possible is called ergonomic prerequisite.
Every system must exist as a 100% functionality before some form of control surface is introduced. Every control surface must be human oriented. In car UX, the main function is driving and everything that reduces risk of removing focus from the road is good UX, everything that creates distraction and complication is not only bad UX - it is dangerous.
In car touch interfaces are exciting for people who hate driving and love their phones. This is the main selling point. It is the result of marketing driven design and enormous amount of production savings (when removing physical controls for dedicated functions) as a business motivational vector.
This will not work long-term. The Idea of vehicles as a SaaS platform also will not work. This will be CX hell and people will reject it.