Stocks - The Best Buys 2

[bluejays]

Originally Posted by Fuzz

Waymo is largely regarded as the leader here, but Chinese companies are making more progress lately. Tesla's main fault is that they've decided to rely on one type of sensor that is cheap. Their logic is that humans have eyes, and drive with eyes only, so a computer should be able to as well. I find this premise entirely flawed, least of which is that humans have brains. So we have eyes and brains. A Tesla has fixed position cameras and a computer. That computer is trained on human drivers who are imperfect. This computer has no knowledge of how a child holding a ball will behave, or a dog, or a cyclist, or any other obstacle on the road. It can only infer actions from it's training set, which will always be incomplete, since every drive you are encountering something novel. Tesla thinks more miles and training can overcome this fundamental problem that can't be solved without AGI.


The other issue with just cameras is we don't just drive with our eyes. We drive with our ears, the feeling of the road, the experience and memory of situations. We also drive on inference when we can't see things, like when a truck passes you in a snow storm. Why wouldn't you want extra sensors for situations like that? They got rid of ultrasonic parking sensors years ago, and the vision only replacement is largely regarded as garbage.



You can peruse the Tesla forums and see just how far they have to go as people describe the vehicle crossing yellow lines, not stopping for red lights, stopping for green lights, phantom breaking, inability to read road signs, and on and on. They have a long way to go, but I fundamentally feel they can't get their with the tech they have. And they've proved that every year since 2016 when Musk has said it's coming by the end of the year. I'm not sure what more evidence is needed that they are nowhere close than what is already out there.


The most accurate description is that it works incredibly well 95% of the time, other than the 5% it's trying to kill you.

Thanks for the explanation - all of it I had no idea. Question for you though, I remember years ago LIDAR was the way to go but was too expensive, so most companies did not bother. Regarding AI, now that we have chips that can help with the learning, aren't we on the cusp of having this stuff figured out, where before it was a challenge?

[Fuzz]

LiAR has come down in cost, probably $1000 now. But it's also the other sensors, like radar and sonar. That help overcome the deficit in thinking.

And no, we aren't anywhere close to AGI. There are fundamental limits tot he current methods being used. Adding more processing power and more training data was thought to be the eventual solution, but the reality is that is just a more refined model that still cannot reason. Current AI has great knowledge, but no wisdom. It's a fancy parlour trick, a prediction engine. For some things, this works well, or good enough. But I don't beleive that technique gets you to self driving, beyond the easy stuff.

Nobody really knows when or if we will achieve AGI. We are in no way close, or else it's just one huge discovery from existing. That huge discovery could be tomorrow, or never. You will know it, because the world will fundamentally change forever.

[bluejays]

Originally Posted by Fuzz

LiAR has come down in cost, probably $1000 now. But it's also the other sensors, like radar and sonar. That help overcome the deficit in thinking.

And no, we aren't anywhere close to AGI. There are fundamental limits tot he current methods being used. Adding more processing power and more training data was thought to be the eventual solution, but the reality is that is just a more refined model that still cannot reason. Current AI has great knowledge, but no wisdom. It's a fancy parlour trick, a prediction engine. For some things, this works well, or good enough. But I don't beleive that technique gets you to self driving, beyond the easy stuff.

Nobody really knows when or if we will achieve AGI. We are in no way close, or else it's just one huge discovery from existing. That huge discovery could be tomorrow, or never. You will know it, because the world will fundamentally change forever.

Thanks. I still may be a little slow admittedly, but what are we fundamentally expecting to change with self driving as it stands today? Say the machine learning today learns that if an upcoming bicyclists' left leg twitches they're likely to fall into the cars' path. The reaction the car should make is to swerve the left to avoid (assuming no car beside). Wouldn't that generally be good enough? What will AGI resolve? I'm still struggling to understand where we need that 1% over the top leap?

[Fuzz]

Originally Posted by bluejays

Thanks. I still may be a little slow admittedly, but what are we fundamentally expecting to change with self driving as it stands today? Say the machine learning today learns that if an upcoming bicyclists' left leg twitches they're likely to fall into the cars' path. The reaction the car should make is to swerve the left to avoid (assuming no car beside). Wouldn't that generally be good enough? What will AGI resolve? I'm still struggling to understand where we need that 1% over the top leap?

Because, as I mentioned, Tesla vision AI is trained on what it knows, and what it has seen. In theory it knows a stop sign means stop because it was trained for that. The problem is that it requires may iterations of each unique case to make a decision on what to do. So what does it do when it sees something novel, a situation it isn't trained for? It fails. It may fail successfully, or it may fail catastrophically. The issue is that as an autonomous vehicle, deployed to millions of vehicles, is that these failures can not happen because someone needs to be liable. We aren't going to accept a collision rate anything less than near perfect. The minute a Tesla creams a kid in a driveway or parking lot is going to be the end of it.



To overcome these flaws, other companies have several sensors, and in my opinion, more robust programing. I think they can get safe enough with these methods, at least in areas with no winter, which is an entire other area of problems.



AGI would bring a human like brain to a vision only vehicle. That brain would be able to reason that a stop sign advertisement on the back of a bus doesn't mean stop. Or that it's Halloween, so you will see some weird things, kids everywhere etc. Or the low sun on the horizon isn't a yellow light. Or that flashing yellow at this particular spot should be treated as a stop sign. Or in this city, if you don't rush the yellow, you WILL get rear ended. And it's not that you can't program these things in, but that there are just too many, it's known as the long tail. And they are constantly changing, such as pedestrians with cell phones. Or headphones on. So AGI will provide the reasoning that allows it to solve for unique issues on the spot, vs pre-decided actions.

[GGG]

Originally Posted by Fuzz

Because, as I mentioned, Tesla vision AI is trained on what it knows, and what it has seen. In theory it knows a stop sign means stop because it was trained for that. The problem is that it requires may iterations of each unique case to make a decision on what to do. So what does it do when it sees something novel, a situation it isn't trained for? It fails. It may fail successfully, or it may fail catastrophically. The issue is that as an autonomous vehicle, deployed to millions of vehicles, is that these failures can not happen because someone needs to be liable. We aren't going to accept a collision rate anything less than near perfect. The minute a Tesla creams a kid in a driveway or parking lot is going to be the end of it.



To overcome these flaws, other companies have several sensors, and in my opinion, more robust programing. I think they can get safe enough with these methods, at least in areas with no winter, which is an entire other area of problems.



AGI would bring a human like brain to a vision only vehicle. That brain would be able to reason that a stop sign advertisement on the back of a bus doesn't mean stop. Or that it's Halloween, so you will see some weird things, kids everywhere etc. Or the low sun on the horizon isn't a yellow light. Or that flashing yellow at this particular spot should be treated as a stop sign. Or in this city, if you don't rush the yellow, you WILL get rear ended. And it's not that you can't program these things in, but that there are just too many, it's known as the long tail. And they are constantly changing, such as pedestrians with cell phones. Or headphones on. So AGI will provide the reasoning that allows it to solve for unique issues on the spot, vs pre-decided actions.

To the bolder this is a failure of people. We currently have insurance because people aren’t perfect. There is no reason self driving should be held to a higher standard. Essentially accident rates should dictate the cost of the insurance to the owner of the self driving vehicle. Smoking a kid in a driveway or mowing down an old lady are perfectly acceptable consequences of driving. I don’t think AGI is required given the success of Waymo.

I think we have seen how far away Tesla self driving is when you compare to the kit on Waymo to the kit on Tesla.

[Fuzz]

Originally Posted by GGG

To the bolder this is a failure of people. We currently have insurance because people aren’t perfect. There is no reason self driving should be held to a higher standard. Essentially accident rates should dictate the cost of the insurance to the owner of the self driving vehicle. Smoking a kid in a driveway or mowing down an old lady are perfectly acceptable consequences of driving. I don’t think AGI is required given the success of Waymo.

I think we have seen how far away Tesla self driving is when you compare to the kit on Waymo to the kit on Tesla.

They are acceptable consequences of humans driving. I suspect human psychology is going to find those types of situations unacceptable when a computer is in charge.

And I think you agree with me, I think using other sensors allows companies like Waymo to be good enough without AGI. Musk has always said humans have eyes only, and can be taught to drive, so they can teach their car to drive with vision only. He neglects the reality that humans think much differently than computers. Other sensors allow for much less inferring about the world around you, because you can decide with more certainty what an object is(and it's properties like direction, speed, etc) when you have vision, radar and lidar. So yes, we already have self driving vehicles safe enough for deloyment with Waymo. No one has demonstrated this with vision only.

I'd kinda half love to see what would happen if Tesla just said all their cars were autonomous now, and sent them on their way. This can happen a long way from wherever I am.

[PaperBagger'14]

Originally Posted by Fuzz

Waymo is largely regarded as the leader here, but Chinese companies are making more progress lately. Tesla's main fault is that they've decided to rely on one type of sensor that is cheap. Their logic is that humans have eyes, and drive with eyes only, so a computer should be able to as well. I find this premise entirely flawed, least of which is that humans have brains. So we have eyes and brains. A Tesla has fixed position cameras and a computer. That computer is trained on human drivers who are imperfect. This computer has no knowledge of how a child holding a ball will behave, or a dog, or a cyclist, or any other obstacle on the road. It can only infer actions from it's training set, which will always be incomplete, since every drive you are encountering something novel. Tesla thinks more miles and training can overcome this fundamental problem that can't be solved without AGI.


The other issue with just cameras is we don't just drive with our eyes. We drive with our ears, the feeling of the road, the experience and memory of situations. We also drive on inference when we can't see things, like when a truck passes you in a snow storm. Why wouldn't you want extra sensors for situations like that? They got rid of ultrasonic parking sensors years ago, and the vision only replacement is largely regarded as garbage.



You can peruse the Tesla forums and see just how far they have to go as people describe the vehicle crossing yellow lines, not stopping for red lights, stopping for green lights, phantom breaking, inability to read road signs, and on and on. They have a long way to go, but I fundamentally feel they can't get their with the tech they have. And they've proved that every year since 2016 when Musk has said it's coming by the end of the year. I'm not sure what more evidence is needed that they are nowhere close than what is already out there.


The most accurate description is that it works incredibly well 95% of the time, other than the 5% it's trying to kill you.

Yeah I think that’s a false statement given the drivers on the road. I’m convinced most have one or the other, but not both.

[GGG]

Originally Posted by Fuzz

They are acceptable consequences of humans driving. I suspect human psychology is going to find those types of situations unacceptable when a computer is in charge.

And I think you agree with me, I think using other sensors allows companies like Waymo to be good enough without AGI. Musk has always said humans have eyes only, and can be taught to drive, so they can teach their car to drive with vision only. He neglects the reality that humans think much differently than computers. Other sensors allow for much less inferring about the world around you, because you can decide with more certainty what an object is(and it's properties like direction, speed, etc) when you have vision, radar and lidar. So yes, we already have self driving vehicles safe enough for deloyment with Waymo. No one has demonstrated this with vision only.

I'd kinda half love to see what would happen if Tesla just said all their cars were autonomous now, and sent them on their way. This can happen a long way from wherever I am.

Yeah i generally agree with you. I missed your first post on the topic where you talked about Waymo. I just disagree with applying a perfect standard rather than a human standard.

I’m hoping the big dollars of Google and Tesla are able to push it to be an insurance risk and not a product liability question. We will have millions of unnecessary humanif perfect is the requirement.

Anyone know how Waymo is insured? If I were Tesla I’d pursue some kind of insurance arrangement where and individual could pay X amount in insurance for auto drive. Very quickly we’d discover if it’s good enough.

[Fuzz]

Originally Posted by GGG

Yeah i generally agree with you. I missed your first post on the topic where you talked about Waymo. I just disagree with applying a perfect standard rather than a human standard.

I’m hoping the big dollars of Google and Tesla are able to push it to be an insurance risk and not a product liability question. We will have millions of unnecessary humanif perfect is the requirement.

Anyone know how Waymo is insured? If I were Tesla I’d pursue some kind of insurance arrangement where and individual could pay X amount in insurance for auto drive. Very quickly we’d discover if it’s good enough.

Tesla has their own insurance they sell to customers, so it wouldn't be a huge step. But the fact they haven't done it should tell you the current confidence they have in that being a success. Reminder, Musk has been saying they are close since 2016.

[bluejays]

Question for those in the know. I don't follow Bell, but just noticed today the dividend yield is at a whopping 11.6%, currently. I don't necessarily have an investment strategy, but the yield seems great for the long term.

• Why has the stock dropped so much?

• Is there a risk of the dividend being cut? Can't say I chase yield but this seems very high for a "blue chip" stock

• What are the chances at the appreciation of the stock? Or will this thing likely continue downward?

Again, not a company I follow. Thanks.

[Slava]

Originally Posted by bluejays

Question for those in the know. I don't follow Bell, but just noticed today the dividend yield is at a whopping 11.6%, currently. I don't necessarily have an investment strategy, but the yield seems great for the long term.

• Why has the stock dropped so much?

• Is there a risk of the dividend being cut? Can't say I chase yield but this seems very high for a "blue chip" stock

• What are the chances at the appreciation of the stock? Or will this thing likely continue downward?

Again, not a company I follow. Thanks.

Well, not investment advice, but their latest quarterly dividend exceeded their quarterly earnings. That’s not a great recipe for success! And yeah, it’s a very high dividend rate, because the stock has been brutal.

The easiest answer for why the shares have done so poorly is high debt load and lower earnings. The dividend payout just adds to that issue, as it’s clearly a huge expense. A piece of their problem is that a dividend cut would also hurt the shares (at least in the short term).

[bluejays]

Got it. I guess I thought the Canadian Telecoms had a monopoly, and with the encroaching into the healthcare spaces, etc., I thought they'd be well positioned and protected, much like the chartered banks. I guess not. Thanks Slava. I guess I'll hold off as it sounds like the dividend has a real possibility of being cut.

[Slava]

Originally Posted by bluejays

Got it. I guess I thought the Canadian Telecoms had a monopoly, and with the encroaching into the healthcare spaces, etc., I thought they'd be well positioned and protected, much like the chartered banks. I guess not. Thanks Slava. I guess I'll hold off as it sounds like the dividend has a real possibility of being cut.

To be fair, I haven’t really analyzed Bell and don’t own it. But I think their dividend payout is like 140% at this point, and that’s only sustainable for so long. Eventually they’ll either cut the dividend, earn more so that’s no longer an issue or borrow money to keeping the dividend. (Or a combination of those three).

I am personally not someone who cares about dividends, so seeing Bell with the high yield doesn’t really get me too excited.

[bizaro86]

Originally Posted by Slava

To be fair, I haven’t really analyzed Bell and don’t own it. But I think their dividend payout is like 140% at this point, and that’s only sustainable for so long. Eventually they’ll either cut the dividend, earn more so that’s no longer an issue or borrow money to keeping the dividend. (Or a combination of those three).

I am personally not someone who cares about dividends, so seeing Bell with the high yield doesn’t really get me too excited.

Aside from the fact they have been borrowing money to pay their dividend and have thus racked up huge debt, the market also hates their go-forward strategy. They sold their share of the Leafs/Raptors to Rogers for a big pile of money. Since those assets are valuable but don't kick out much cash this was regarded as a way to reduce debt and right the ship.

Then almost immediately they spent all the money plus a bunch more buying us Fibre assets at a high price. They don't have US cellphone assets, so there's no synergy or cross-selling opportunities, and since they have minimal US operations they probably can't cut any costs. But at least they paid a very high multiple of earnings/EBITDA...

[GranteedEV]

I just bought about $6400 worth of Green Thumb Industries, current price is $10.95 CAD. Pretty volatile stock, definitely not going to recommend it to anyone who hasn't done their own due diligence, but I like that I could buy it on a canadian exchange (when USD is very high) at what feels like a low price. My hope is to either swing it if it sees a huge random gain in the next 90 days (simply because I put a +20% 90 day limit sell on it) or forget about it and hold it long term until US regulations allow it to be traded on american exchanges (it's clearly their goal, otherwise it would be on the TSX which has stricter rules for that sort of thing). Once this hits the NYSE or NASDAQ it's going to see some significant gains in my opinion, because US-focused investors will jump on it. It has a market cap of 2.75B right now but i see no reason it can't 15X that.


Beyond that, if I didn't just spend 6400 on weed stocks, the trio of quantum stocks in this article:

https://www.google.com/amp/s/www.mar...check-7fbdf443

look like a nice buy for a long term hold given how dramatic that fall was... i hate calling anything "the future" but quantum computing value to solving real world problems like medicine alone mean for me, that "reality check" is good news. Of course 15-20 years is a long time to hold so ymmv

[Fuzz]

Quantum computers are in use right now. I don't recall the name of the company, but Canada has one company that is using them.

[Dan02]

Originally Posted by Fuzz

Musk has always said humans have eyes only, and can be taught to drive, so they can teach their car to drive with vision only.

He's 100% right humans are driving using only vision, the part he's forgetting is humans ####ing suck at driving, 1.35 million people die PER YEAR on the roads.

Computers definitely have the advantage of not being distracted/impared and being able to see 360 degrees around them at all times, but maybe self driving cars should take advantage of additional sensors humans can't use.

His excuse is the sensors are expensive, but just like batteries, rockets, starlink satelites, all things he has experience with, scale improves cost.

[Fuzz]

Computers have an attention advantage, but lack intuition and understanding.


You are correct that humans suck and are getting worse at driving, but a Tesla is still far worse. The most recent real data I can find is from October, requiring and average of one critical disengagement per 13 miles. This is obviously not safer. I think I had read it is closer to 17000 miles for Waymo, which they consider safer than humans. So Tesla has a long way to go, has been working on it for a decade, and is still literally thousands of miles from success. I don't see a trajectory that gets them there with their tech, unless it's all just a grift.

[GomerPile]

Was sitting a lot in cash, couple weeks ago I moved it all in US Stocks. Mostly big names, seems like the Trump bump is in full effect. Wondering what everyones thoughts are for length of bump?

My original theory was to sell some of it once the tariffs came in and us the high USD to buy back into Canada. Mostly the plan seems to be working...

[Fuzz]

In this #### show? If anyone tells you they know what to do, they do not. You can not predict chaos. You can get lucky and guess correctly, but I don't think you can predict the right move in this environment with any confidence.