With his level of promise to reality, we might have a robot that can carry a couple boxes across a room without falling over by the end of the decade.
If Elon Musk's claims made during Tesla's shareholder meeting this week are accurate, get ready for that humanoid robot he promised, some self-driving software update, and an overhauled Cybertruck. Then again, this is Elon we're talking about. The tech tycoon's portion of Thursday's meeting consisted of a Tesla corporate …
I know it was a joke but Auto park has mostly worked reliably on lots of models of cars for years. My 2013 Merc has it. Though to be fair last time I used it parallel parked perfectly lengthways across 3 spaces. And remembering how to turn it on takes more mental effort than parking myself.
The other 10% of the time YOU HAVE TO DO IT. You don't just let the damn car CRASH unless you are a particularly dumb person who values their "principles"* over their life.
(* "Dammit, it says Full Self-Drive so by God it ought to do it! I'll be damned if I'll turn the steering wheel to save my own goddamn life when I paid a lotta money to have car do it!" is actually too long and profane to go on most headstones but otherwise that's what it would say)
I've got ENOUGH "Insider Information" about Optimus that I am confident that our NCA (North Canadian Aerospace) Electro-Dynamic Wireline Musculature (EDWM) droids "Tyrone" and "Ensoniqa" quite outperform Optimus in mobility, smoothness and gracefulness, strength, agility and environmental awareness.
When WE at NCA build droids, we REALLY build droids! While they are NOT LIGHT, weighing 500 lbs (200 kg) each due to their all-Titanium construction along with realistic looking (and heavy!) silicone skin they can lift 1000 lbs (400 kg) and run 40 kmh (24 mph) quite outperforming any human!
AND YES! I did code the 2D-XY/3D-XYZ SOBEL/CANNY fully-autonomous vision recognition system for it!
AND YES! It is using a few of our 128-bits wide combined CPU/GPU/DSP/Vector Array Processors as its brain. The power supply is a bit iffy since our Aluminum-Sulfur batteries are a tad heavy and only lets it run for about an hour (60 mins) at full lifting and running capability.
YES! it's a bit dumb since even Two+ PetaFLOPS (i.e. each of the 4 installed chips are 575 TeraFLOPS of computational horsepower) is STILL not enough to match human environmental awareness abilities! Most of the CPU horsepower is dedicated just to machine vision, localized hearing and English-language understanding. Only a fraction of that computational horsepower is needed for 3D-XYZ control over environmental contact with ground surfaces and finger/hand/arm and leg/knee/ankle/feet/toe coordination and motor control for balance, walking, running, obstacle avoidance and lifting purposes.
Wireline-based musculature which uses grouped conductive polymer strands that CONTRACT in concert when a specific voltage/current is applied and relaxes when it is turned off, allows for great MIMICRY of human musculature and is FAR MORE POWERFUL than ANY human muscle group. Since it works like the human musuclo-skeletal system, it also LOOKS and PERFORMS just like you and me!
AND YES! "Ensoniqa" is definitely "Hot Looking"! --- For the ladies, "Tyrone" is also stacked like a boxer!
We COULD make them fully functional in every way possible but that would probably mean the death of the human species due to lack of sexual competition! aka See Futurama Cartoon Skit "Don't Date Robots"!
While all of our staff is thinking it .... AND..... there is a TACIT UNDERSTANDING that we will "Eventually Do It", android/gynoid sexual competency will have to come at a later date! he he he!
Anyways, these bots have been in LONGER DEVELOPMENT than even our super-CPU chips (20 years versus 12 years now!) and only until our Aluminum-Sulfur battery discoveries were we able to make these 'Bots practical for non-tethered power and comms cable use.
I SHOULD NOTE THOUGH! While they are high-performance physically, Tyrone and Ensoniqa have the intellect of maybe a Racoon or House Cat. It will take a MASSIVELY POWERFUL stacked block of Gallium Nitride brain-blocks to get to average IQ-100 human level intellect in a small space. We do HAVE actual 160+ IQ superintelligence-level WBE (Whole Brain Emulation) supercomputers BUT they are literally inside of and nearly the size of a large mountain!
My "private demo" of Optimus didn't exactly "excite me" all that much! However, from a general public's viewpoint IT IS VERY VERY GOOD and MANY OF YOU will be SHOCKED at its agility and fluidity and it will AMAZE the global media and general public BUT if I put Optimus Side-by-Side with Ensoniqa and Tyrone AND any modern male and female human, Optimus PALES in ability and looks compared to our NCA-built humanoid systems!
Now You Know!
We are Canadian. We make superchips, supercomputers, autonomous robots, cameras, batteries and spacecraft. We make BETTER robots than Telsa. We make them out of Titanium, Silicone Rubber and conductive polymer muscles. They are PHYSICALLY POWERFUL and SMOKIN' HOT looking! They CAN and WILL replace all human in almost all jobs in the future!
Does THAT work for you?
They CAN and WILL replace all human in almost all jobs in the future!
Good thing I work in the present.
Props on some quality kookery, though. I was getting a bit tired of Geller's "I invented AI, fools!" posts, so a little variety is appreciated. Sorry, I mean a little VARIETY is APPRECIATED, THANKS!
Try this link to find some OPEN SOURCE DISCLOSURES on new camera and image sensor hardware!
North Canadian Aerospace
Scandium Image Sensing Inventions Open Source Disclosure Under GPL3 Aug 23 2021:
(softwall that needs account and login to see rest of docs past first 3 pages)
You can also goto CanonRumors to look for Scandium Image Sensor in the 3rd Party Products forums to see the full docs. These were disclosed ON PURPOSE TO PREVENT PATENTS being filed by other parties.
Does that help?
I must admit, that term gigafactory bugs me too. Just how big is a Tesla "gigafactory" compared to a run of the mill car factory? Does it cover more space? Is it more vertically integrated with more on-site manufacturing? Or is it just a new "cool" name for an assembly plant with no actual manufacturing capability?
Tesla has not even come close to reliably solving one relatively simple task," Marcus said, referring to the years of work on self-driving that Tesla has yet to perfect though...
I agree with him that it's preposterous that the robot meets all the bollocks that Elon claims, but I have to disagree that self driving is a" relatively simple task".
I don't think anyone in the field of optical recognition or self driving would ever think it was a simple task. It's a very difficult problem, even more so when you take it out of the lab. All the more reason Tesla should get a locking for such blatantly false marketing claims as "Autopilot" and "FSD"...
Self driving is absolutely not a "relatively simple task", it is a very simple task compared to developing a robot that can replace human manual labor.
Musk clearly has no idea about the engineering effort involved in doing the things he promises. He just assumes if he hires enough PhDs and engineers and puts them in a room together, magic will happen. That might work for mostly already solved problems that require only basic physics and equations that are easy to calculate like launching a rocket and having it land upright, or are pure engineering problems that also have been done before like replacing an engine and transmission with electric motors and batteries.
But for breaking new ground in problems that can't be reduced to equations and are way beyond ordinary engineering requiring many advances in the state of the art, like self driving cars, mind reading headbands and human analog robots, he seems to think solving them will be equally easy. In those domains he's simply got no clue so he's all promise and no follow through.
What I wonder is whether he's being told how difficult the problems are and he chooses to ignore the people he's tasked with actually solving the problems, or whether they simply tell him what he wants to hear and he's not smart enough to realize they're patronizing him.
> Self driving is absolutely not a "relatively simple task", it is a very simple task compared to developing a robot that can replace human manual labor.
So you *do* think that self-driving absolutely *is* a relatively simple task, compared to a humanoid robot! Which is the comparison that Gary Marcus made.
Yet you say you don't, then you say you do, but you don't.. Norman, co-ordinate (clunk)
Having seen prototypes of OPTIMUS IN PERSON (i.e. I got a private and secret demo from a personal "friend"), I can quite assure you that his engineers are a lot further along than you think! Optimus is getting near Boston Dynamics Atlas in a lot more compact and quieter package. It's actually pretty shocking how far he has come.
Our own company (NCA -- North Canadian Aerospace) has been doing humanoid robotics for ALMOST 25 YEARS NOW (1998) and it took our company spending over five-and-a-half-BILLION DOLLARS making a 128-bits wide YottaFLOP+ supercomputer before we were able to do the human body form's motion modelling and simulation at a high enough resolution that we could put that dataset into an android and gynoid format and STILL fit all that motion data and synthetic vision/audio system technology into four 575 TeraFLOP chips that could actually physically fit into a normal size human-looking head. It also took our company's robotics division TWO DECADES to find the right class of super-strong conductive polymers in order to EMULATE and MATCH the fluidity and smoothness of human musculature and body movement.
Robotics is NO EASY TASK and I actually APPLAUD Tesla for coming as far as they have within just the last THREE YEARS!
YES! Tesla Optimus is NOWHERE NEAR the capabilities, looks and fluidity of our in-house designed and built "Tyrone" and "Ensoniqa" fully autonomous android/gynoid systems BUT it is A LOT BETTER than most people think and imagine, plus they SPENT NOWHERE NEAR what we did over the last 25 years! They have come a LOT further in a lot shorter timeframe than we did. We are MUCH MUCH FURTHER AHEAD ahead due to our in-house supercomputers being able to model the human form with near-perfect duplication BUT Tesla is doing a LOT BETTER than most people think!
YOU WILL BE VERY SURPRISED by SEPTEMBER's/OCTOBER's Tesla Optimus Demonstration!
From a long-time computer science trained vision coder perspective, Optimus is actually pretty damn impressive for what it is! It's a LOT better and farther along than I thought it would be!
Gary Marcus is wrong on both counts, driving on public roads is a hard to intractable problem, Task based robotics, not so much. General robotic labor would be harder, but why would you torture yourself? A bigger problem is that humanoid robots are niche, verging on pointless.
With the amount of time and expert labor to program a squad of humanoid robots to correctly execute their tasks, you could just as readily deploy automated tools and stations that are on carts or straight bolted down. No need for legs and all of that complicated programming. Are human arms really the best fit? If they are great, otherwise just mount the power drill it will be wielding onto something with the right number of axes.
Then you suddenly realize we had all this stuff decades ago. These people keep trying to sell us on the idea we need humanoid robots to flip burgers. A robotic burger grill is going to look like a smaller version of a pizza oven. Again, they already exist, you just need to tie the tools and stations into an assembly line.
This isn't the jetsons.
But that is the point being made. This would be a generalised robot, supposedly capable of replacing a human, not a specialised robot, which as you said would then a specifically designed automated system would be far faster and cheaper to make.
A generalised robot capable of replacing a human is far more complex than driving, as, technically, if its a generalised robot, capable of replacing a human, it should be able to drive also :)
And then I realised that I know nothing about modern vehicle terminology and found out that the above two comments should have referred to the Tesla "Semi" not the "Cybertruck".
Hmm, the charabanc is parked a bit away; shall need the velocipede to get there. Toodle-pip.
Honda introduced Asimo some twenty two years ago, and Boston Dynamics introduce a new robot on a monthly basis, it wouldn't surprise me in the slightest if Musk unveils a "working" humanoid form robot this year.
However, the definition of "working" will depend on who you're talking to. The fans will insist it (like FSD) is just one software update away from being a real boy.. sorry being a functional human assistant. The cynics will point out that so far autonomy has escaped all researchers, and the Honda and Dynamics machines are expensive, complex and require human direction to perform most tasks.
As such, practical uses of this class of machine tend to be extremely limited. Like the mythical robotaxi, the futurologists get excited about a future that just doesn't bear much scrutiny.
Yep, at best it will pick up a box and carry it across the stage and put it down, and his acolytes will cheer wildly, and Tesla stock will jump after he tweets something stupid claiming there will be a million robots in two years replacing people performing all sorts of dangerous work.
Like the zero Tesla robotaxis that will be on the road earning $100K for their owners each year by the end of this decade (when he promised that in 2020) we'll see zero Tesla robots replacing a single human in any task in the same timeframe. Or ever. At most you'll be able to tell them to grab you a beer from the fridge, and even that will take him years to manage.
After having seen a "Private Demo" (unapproved by Elon of course!), AND noting that I basically have 30+ years of Synthetic Vision Systems coding experience, I am quite surprised at HOW FAR ALONG Tesla is after only THREE YEARS of in-house development. It is my opinion that teh Tesla Optimus android will start surpassing Boston Dynamics Atlas as soon as the next software update. It's smaller, smoother and quieter than Atlas and matches its agility.
The September/October demo will be limited in scope BUT it does showcase that Tesla Engineers KNOW what they are doing and that Optimus will be a viable GENERAL PURPOSE REPLACEMENT for MANY labour intensive service jobs within as short a time frame as 5 years!
Sorry kiddos! I've been doing vision a LOOOOOOONG TIME! Optimus is a LOT BETTER and farther ahead than you think!
Amazon Warehouse workers, UPS, Fedex and even Starbucks and McDonalds workers are pretty much DOOMED! By 2027, you will start seeing Optimus (or NCA) bots in multiple business settings. General warehouse order picking, sorting, shipping/receiving, facilities cleaning and backroom food services are THE FIRST JOBS that will be replaced by these systems. After that it will be hotel/retail backroom jobs, room cleaning, dishwashing, and lower-end cooking/baking services. Delivery persons and front desk CSR work and coffee baristas, food servers.hostesses will be soon after. It really ONLY is just a battery problem that is holding people back from going all out on robots being EVERYWHERE in EVERY JOB!
(Sorry! But I just can't help this big dig though: Optimus is DEFINITELY NOT as agile, smart, fast or as "Hot Looking" as the in-house NCA-built Tyrone android and Ensoniqa gynoid systems! --- We are literally MILES AHEAD of Boston Dynamics Atlas, Festo, Honda Asimo, Sony, Samsung, Hyundai, Tesla Optimus, etc, etc. --- Ours at least LOOK and ACT human due to the onboard FOUR of our 575 TeraFLOPS 128-bits wide super-chips and the very realistic looking silicone rubber skins!)
Wait for the presentation : The Optimus robot transforms into a cybertruck.
Either that, or this year the actor in the bodysuit will have an aerial stuck on their head.
I'm sort of suspecting lots of other mostly ridiculous ideas, full of wild overpromising, which he "sadly" will end up being unable to bestow upon the human race, due to the financial meltdown of his idiotic twitter takeover fiasco.
There was also a semi that was supposed to be on the road in the thousands long before today that not a peep has been heard from.
I'm sure there will be some new bullshit on stage, because his believers are too stupid to remember all the failed promises of the past and will drive up Tesla's stock price based on the wild claims he makes for new products he claims to be working on. That will give him enough ready cash to buy Twitter at the agreed upon price as the court is likely to force him to do.
Didn't your mother teach you that swearing was not a good thing ?
Honestly, with all the stupid things you've said for the past twenty years, one would think you'd have learned to shut up by now.
But no, obviously you don't learn a thing.
By the way, how's that Hyperloop going ?
Oh, yeah, it's not.
For a couple of centuries after the widespread adoption of the printing press, 'social norms' involved burning unpopular women as witches.
I'm sure if you think about it, it won't take long for you to find other examples in history of social norms being different to our own.
So it would appear that conforming to social norms is not in itself a route to goodness.
Arguments based upon first principles, on the other hand, I can attempt to engage in. Whilst trying to avoid the hubris of believing that I'm not shaped by the social norms that went before me. That why Matt Groening and Kurt Vonnegut identify with the Unitarians.
Tesla are no longer alone in the luxury end of the EV market, for example the Jaguar I-Pace base model is a bit dearer than the basic Tesla, but the top of the range Jags are a lot cheaper than the top of the range Teslas.
Can't see Tesla maintaining the proportion of the market they've enjoyed over the last few years.
As for the robot, if connected to a computer through a power and control umbilical then no real problem, but a self powered autonomous humanoid robot that will run for more than 30 minutes on a charge, no chance for at least 10 years.
I drive around Vancouver, British Columbia, Canada every workday and I am almost never out of sight of at least one Tesla. Sometimes 2 or 3 at a time. And parades of Amazon vans coming out of several massive warehouses. I am not at all pessimistic about seeing them all driverless this decade. If civilization survives climate change etc for that long.
But how many of the non-Tesla cars are EVs or hybrids? It can be tricky to tell from a distance.
Where I work we've put some public facing charge points up and while Tesla is probably the single most common manufacturer we see represented, they are by far outnumbered by non-Teslas and the ratio is shifting away from Tesla every month. Most of the increase is at the less luxurious end of the market with Nissans, light commercials and similar sized cars although the Jags are becoming more common.
The market penetration depends on charging facilities, anyone who can afford £50k+ for a car will probably live in a house with off-road parking so can install a charger but a large proportion of people on lower incomes do not have anywhere to place a charger so we need far more public chargers.
Yeah, I live in an end terrace, so cannot guarantee getting parked outside my house, so a pure EV is out of the question. I recently test drove a PHEV, so when we move, I could utilise the plugin aspect and get cheap travel for local journeys. On public charging points, the local supermarket has charging points, they were free to use at one point, but there's no infomration on their web site about whether they are free, or chargeable, or what the spec is,.. so they've gone to the trouble of putting them in, but aren't really poromoting them, which seems odd to me.
You can probably charge off a standard outlet, and in most cases you will have to. There is little incentive for installing the non-fast DC charge ports, so if you have a L2 plug, best of luck, there will be 4 cars for every plug and that will keep getting worse as they aren't adding new plugs, but the number of cars is going up.
The economics of pay to charge for L2 aren't attractive, as the cars need to stay docked longer, and take less power, so the margin is terrible, Some places like malls and markets dropped a couple of free chargers in the back of their lots, but there aren't nearly enough. And there is no national level plan to ensure that new charge points support the (supposedly) national standard L2 plug. So if you have an non-dc charge EV or PHEV you are going to need to charge at home or work(if you are very very lucky).
Even then, the DC charge format will probably change again in a couple of years, leaving the same issues to play out all over again. They should have either mandated DC charge points to have a 240v plug point available, or made it so you could drop in an adapter to the vehicle/charge cable.
This post has been deleted by its author
Vancouver is FILLED with Teslas and other EVs! I've even seen the Ford F150 Lightning and various BMW EVs. Here in Metro Vancouver area, environmentalism has always been a big thing here. Soon EV users will be able to ENJOY the use of fancy new battery technology super-fast charging and advanced environmentally friendly Megawatt to Terawatt-class mains power production systems!
As long as:
- it's got some stupid 'fun' mode where it dances a jig if you say 'Elon is a god'
- it's very very expensive
- it has no practical use-cases... yet. They're all 'planned' and will be released 'in due course'
- it can prepare coffee (after safety software update)
- it can carry a specific box safely across a mapped room in less than 10 minutes (after safety software update)
- it has a battery that lasts nearly an hour (though it'll actually last 10 minutes; if you want an hour, you'll need to pay extra per month to 'unlock' the battery)
- it has a 90% success rate in walking in a straight line for 10 metres without killing a spectator
- it has a gun (American market only)
- it doesn't really exist
"Tesla had recently manufactured its three millionth car since beginning production in 2012"
And Ford have shipped some 350Million since 1903, so it still amazes me Tesla has a higher market cap. Well, a higher market cap than all the top motor manufacturers combined. Fun bit of history, in early 2000, Palm Computing Inc had a higher market cap than Ford. Remember Palm?
The 125-pound 5'8" robot will allegedly be able to walk at 5 MPH, carry 45 pounds and deadlift 150 pounds, using visual sensors in its head and "human-level" hands.
That's not even a 25kg lift (UK bag of cement weight & also UK male Health & Safety safe lifting weight for men (16 for women, though nobody told the woman at our local merchants who nonchalant carries heavy bags of around as if they are full of feathers ))
... and obviously lots of places flout the minimum weight rules as any parcel delivery driver will tell you - over 25kg packages occur quite often (and packages exceeding 16kg female limit, very often)
So, not going to be a person replacement in many manual handling roles.
When I was a child, all the heating in our flat was done by coal fires. Cooking and hot water was gas, but heating each room was by open coal fires.
Every couple of weeks the coal lorry would stop by, and if you needed coal you'd shout out the window something like "One sack to flat 20"..
These flats were 4 floors high with no lifts.
The coalmen would RUN up the stairs with a 1cwt (112 lb.) sack over their shoulder!
I suspect they never needed to go to the gym...