• New forum software - please post issues in this thread

Artificial Intelligence

aquabliss

Well-known member
Lots of additional SW Engineers, Technical Program Managers and Support staff needed, but this is a much different skill level than the jobs the automation is replacing.
 

nosuchreality

Well-known member
aquabliss said:
Lots of additional SW Engineers, Technical Program Managers and Support staff needed, but this is a much different skill level than the jobs the automation is replacing.

Lol, oh its coming for those too.

BTW, 2030 is closer than the housing bubble pop.
 

usctrojancpa

Well-known member
nosuchreality said:
aquabliss said:
Lots of additional SW Engineers, Technical Program Managers and Support staff needed, but this is a much different skill level than the jobs the automation is replacing.

Lol, oh its coming for those too.

BTW, 2030 is closer than the housing bubble pop.

Let me know when AI will replace real estate agents so I can retire.  :D
 

Kenkoko

New member
USCTrojanCPA said:
nosuchreality said:
aquabliss said:
Lots of additional SW Engineers, Technical Program Managers and Support staff needed, but this is a much different skill level than the jobs the automation is replacing.

Lol, oh its coming for those too.

BTW, 2030 is closer than the housing bubble pop.

Let me know when AI will replace real estate agents so I can retire.  :D

It's already happening.

Big money is coming into real estate. Private capital invested in real estate technology companies increased from $28 million in 2008 to a projected $3 billion in 2018, with almost all of it going to disruptive brokerages, not ad-driven listing-search sites.Over that same period, a boom for real estate, the largest holding company of traditional real estate brokerages, Realogy, lost nearly 60% of its market value, despite five straight years of increasing revenues.https://www.redfin.com/blog/housing-impact-of-2008-financial-crisis/

The realtor lobby also no longer has the absolute stranglehold on our legislature. The National Association of Realtors, one of the largest and wealthiest lobby groups in the U.S., emerged from the recent tax overhaul hobbled and humbled. https://www.wsj.com/articles/realto...ds-get-ready-for-their-next-battle-1515538892
 

usctrojancpa

Well-known member
Kenkoko said:
USCTrojanCPA said:
nosuchreality said:
aquabliss said:
Lots of additional SW Engineers, Technical Program Managers and Support staff needed, but this is a much different skill level than the jobs the automation is replacing.

Lol, oh its coming for those too.

BTW, 2030 is closer than the housing bubble pop.

Let me know when AI will replace real estate agents so I can retire.  :D

It's already happening.

Big money is coming into real estate. Private capital invested in real estate technology companies increased from $28 million in 2008 to a projected $3 billion in 2018, with almost all of it going to disruptive brokerages, not ad-driven listing-search sites.Over that same period, a boom for real estate, the largest holding company of traditional real estate brokerages, Realogy, lost nearly 60% of its market value, despite five straight years of increasing revenues.https://www.redfin.com/blog/housing-impact-of-2008-financial-crisis/

The realtor lobby also no longer has the absolute stranglehold on our legislature. The National Association of Realtors, one of the largest and wealthiest lobby groups in the U.S., emerged from the recent tax overhaul hobbled and humbled. https://www.wsj.com/articles/realto...ds-get-ready-for-their-next-battle-1515538892

Not sure how Redfin, Zillow, Offerpad, or any other of the other new disruptor companies have eliminated the need for humans to be involved in completing a real estate transaction. For all of these companies (especially the public ones), at some point it'll be all about the profits and last I checked they were all losing money.  Innovation and all is great but it has to be profitable.  Zillow's home flipping arm couldn't seem to make money, what's the deal?
https://finance.yahoo.com/news/zillows-stock-plummets-after-q2-earnings-disappoint-143225540.html
 

qwerty

Well-known member
We are in the early stages of real estate technology. I don?t think it will ever eliminate the need for human involvement but the human involvement may not be in the form of agents, probably back end support staff to process the paperwork and low paid people to show the homes. Amazon was not profitable when it started and now it?s a beast. Only time will tell.
 

usctrojancpa

Well-known member
qwerty said:
We are in the early stages of real estate technology. I don?t think it will ever eliminate the need for human involvement but the human involvement may not be in the form of agents, probably back end support staff to process the paperwork and low paid people to show the homes. Amazon was not profitable when it started and now it?s a beast. Only time will tell.

I'm sure the machines will take over with real estate too but I'm be long retired when that happens.  :D
 

Kenkoko

New member
USCTrojanCPA said:
Not sure how Redfin, Zillow, Offerpad, or any other of the other new disruptor companies have eliminated the need for humans to be involved in completing a real estate transaction. For all of these companies (especially the public ones), at some point it'll be all about the profits and last I checked they were all losing money.  Innovation and all is great but it has to be profitable.  Zillow's home flipping arm couldn't seem to make money, what's the deal?
https://finance.yahoo.com/news/zillows-stock-plummets-after-q2-earnings-disappoint-143225540.html

Big money see profit opportunity because the current real estate model is so inefficient. That's why even if Zillow's flipping arm fails, big money will keep coming into this space. Eventually, something will stick.

If you look at real estate in east asia, they are already seeing much more transition away from human agent involvement especially in non-luxury RE. Nowadays it's some college kid wearing uniform and getting paid min wage to open doors.
 

irvinehomeowner

Well-known member
You probably should take your world trend from an leading automaker instead of a lagger. Elon Musk says Tesla will go fully autonomous by end of 2019. I hate to knock Ford but they are really the bottom fedder of the industry. There's a reason why Ford's stock is down 20% since Tesla's IPO and Tesla has gone up more than 10x

https://cleantechnica.com/2019/02/2...-unequivocal-tesla-autopilot-improves-safety/

I like to revisit these threads because time is the true arbiter of what really happens.

First off... fully autonomous driving by Tesla is still in beta... 3 years later.

In fact, Teslas may not have enough hardware to do it:


Secondly, Ford is doing just fine. They actually came out with an EV truck before Elon did... the Cybertruck is still vaporware right now. And their EV sales are increasing:


As I've said before, I may not be as close to AI as Ken, but I know software development and infrastructure... adoption will always be a bottleneck unless a larger group embraces the technology. We are just finally seeing that with EVs but AI will take much longer.

That being said... I like Hey Google much better than Hey Siri. Not sure how Alexa or Cortana are coming along.
 

CalBears96

Well-known member
In my opinion, full self driving capabilities cannot be adopted until there's an infrastructure to control all the cars on the road. So yes, the bottleneck is there until all the car manufacturers embraces the technology. I would love to see it happen, but I'm not optimistic about it.
 

OCtoSV

Active member
In my opinion, full self driving capabilities cannot be adopted until there's an infrastructure to control all the cars on the road. So yes, the bottleneck is there until all the car manufacturers embraces the technology. I would love to see it happen, but I'm not optimistic about it.
Agree - we're a loooong way away. But the autosteer and adative CC on the Tesla are good enough for me.
 

marmott

Active member
I have a ICE car from 2019 with autosteer and adaptive CC (MobileEye solution) that is also pretty reliable. It comes with some safety limitations, it needs one hand on the wheel past 40MPH for example. But I'm curious where have the major advancements been since?

In fact, Teslas may not have enough hardware to do it
Removing the radar/lidar sensors has put even more strain on the computing power because everything goes through vision now and it's much more compute power hungry.
 

OCtoSV

Active member
I have a ICE car from 2019 with autosteer and adaptive CC (MobileEye solution) that is also pretty reliable. It comes with some safety limitations, it needs one hand on the wheel past 40MPH for example. But I'm curious where have the major advancements been since?


Removing the radar/lidar sensors has put even more strain on the computing power because everything goes through vision now and it's much more compute power hungry.
go Turo a new Model 3 for a weekend and see - max Autopilot speed now 85mph. I basically touch it with my finger when the "hold wheel" alert comes on. So precise, and I like the algorithmic braking vs my own for cutting down on excess wear and tear on the brakes. Combined with the large foot wells on the Tesla it makes a long distance drive let alone everyday traffic feel almost effortless. Way beyond any other mfrs system as it's all about the software and hardware, and do you really think the ICE mfrs have software teams of the caliber anywhere close to Tesla's let alone their own silicon design teams? I couldn't imagine getting more vehicle for $60K - the value is off the charts.
 

irvinehomeowner

Well-known member
Telsa admits that the current hardware set up (even before they removed lidar and ultrasonics) isn't enough for FSD.

That brings up the issue with AI, there will always be hardware/infrastructure limitations that will hinder implementation because some applications are not just software-based.

Let's just use camera/Vision that Elon thought was going to be enough for FSD. His claim was that people use their own eyes to drive so a car should also use just cameras in order to enable FSD. But.... Teslas don't have cameras in the forward bumper... so that's a huge blind spot. Additionally... the resolution of the cameras may not be high enough... and lighting conditions also affect the accuracy of those cameras.

So, an AI can probably extrapolate that data based on the limited input it has.. however... the onboard chip won't have enough computing power to do that in real-time (even with the AMD upgrade).

Someone might say "Just leverage the cloud"... but there is latency and coverage issues with that. While you're waiting for Tesla-Skynet to tell the car to swerve because someone not using FSD is going to clip you... it may have already happened.

There is also the human aspect... for true AI to work in this example of automated driving... all cars need to be on a connected system. If even one is manual... that could mess it all up. It really needs to be like Minority Report where once you get on a highway... all cars are controlled by a central computer.
 

CalBears96

Well-known member
There is also the human aspect... for true AI to work in this example of automated driving... all cars need to be on a connected system. If even one is manual... that could mess it all up. It really needs to be like Minority Report where once you get on a highway... all cars are controlled by a central computer.
This is what I meant when I said that there must be an infrastructure to control all the cars on the road for real automated driving to work. One car not connected to the system could mess up everything.
 

irvinehomeowner

Well-known member
It should be noted that there is a difference between automation and AI.

Automation is easier to implement because there should be no variances... AI is all about dealing with infinite variances.

While some aspects of Full Self Driving are automation... the unknown variances are going to be difficult for AI on the scale of a car computer.
 

marmott

Active member
go Turo a new Model 3 for a weekend and see - max Autopilot speed now 85mph. I basically touch it with my finger when the "hold wheel" alert comes on. So precise, and I like the algorithmic braking vs my own for cutting down on excess wear and tear on the brakes. Combined with the large foot wells on the Tesla it makes a long distance drive let alone everyday traffic feel almost effortless. Way beyond any other mfrs system as it's all about the software and hardware, and do you really think the ICE mfrs have software teams of the caliber anywhere close to Tesla's let alone their own silicon design teams? I couldn't imagine getting more vehicle for $60K - the value is off the charts.
I'm still not convince that what you are experiencing on the highway in your $60K M3 cannot be experienced in the $3xK 2023 Bolt EUV with Super Cruise. Again level 2 ADAS systems are not new, my 2019 ICE does lane keep assist, adaptive cruise, lane change assist and all the flavors of emergency breaking and collision avoidance.

I think the compute in their car is overpowered for level 2 driving assistance but under powered for full autonomous (level 4/5). You end up in the situation today where the car can sort of do autonomous but it not good enough at it.
 

Danimal

Active member
It’s nice that Tesla giving $3750 credit for December delivery on M3 & MY. My MYLR VIN just got assigned. It’s Austin built. No USS though. Cant imagine how Elon thinks he can get away with FSD w/o USS.
 
Top