Agreed. I think the best use of this sort of tech is to use both to their strengths. Use AI to go over the record and suggest diagnoses which you have the doctor review after observing the patient.
The other thing is that common issues are common. I have to wonder how much that ultimately biases both the doctor and the LLM. If you diagnose someone that comes in with a runny nose and cough as having the flu you will likely be right most of the time.
Not that crazy. It's about the most basic structure you can make. Hard to make a better wheel.
The closest thing I can think of that's come close to maybe challenging DRAM is HP's memristors but those really didn't pan out (probably too much power consumption).
Pet peeve: stupid analogy seeing how wheels kept being improved throughout the millennia with every new technology. The only thing in common is that it's round.
Similarly, DRAM in any way you see it has been improving to the point of barely being recognizable since the 70s.
That said, DIMMs and the whole bus idea is in dire need of getting a new type of bearing.
> That said, DIMMs and the whole bus idea is in dire need of getting a new type of bearing.
IBM has been using their own memory bus technology for both their POWER and Z machines. IIRC, it’s somewhat reminiscent of CXL, trading latency for bandwidth and size.
Seems like a pretty good analogy as you admit there have been advancements while the basic structure has remained the same. Not sure why you have a pet peeve when it is highly analogous.
The ultimate shape of DRAM is the same, the main thing that's changed is the materials and techniques to produce it. Making it very impressive, but none the less completely recognizable by someone who was familiar with DRAM in the 70s.
The wheel is the same. Pluck someone from 1000 years ago and they'll be able to correctly identify a modern wheel even though they've never seen any of the composites that go into it. The function of the wheel is identical to how they used it.
You're right—the wheels on the Boeing 737 are, although made of forged aluminum or magnesium to withstand extreme force and heat, pretty much the same shape and operate in the same way as the Ljubljana Marshes Wheel of 3150 BCE.
Then again, flight itself has obviated—or, rather, introduced—many transit workloads that could be performed by wheeled vehicles, and operates on different principles entirely.
I found it fascinating that jetliner tires are designed to blow during extreme braking. An aborted takeoff (say microburst warning goes off), or an emergency landing will heat the brakes and tires to maximum. But there are plugs that will depressurize the tires without them coming apart. That way they don't catastrophically blow and spew tire pieces everywhere to get sucked into jet engines and damaging them.
pretty cool, sort of graceful (or at least planned) degradation of tires without cascading problems.
I'd guess that it'd allow for thinner layers which is ultimately why you can pack in more memory.
And why it's not currently done is likely because it's hard enough to stack when everything is uniform. A small deformity in the first layer will spoil the entire chip.
The voting rights act was a recognition that there's something wrong if a state with 10 congressional representatives and 30% black population ends up with only 1 or even 0 black representatives. That was often done by drawing districts such that the 30% of black voters are diluted by the 70% white voters.
The supreme court has called this "racial gerrymandering" which is a bit rich considering they previously ruled that gerrymandering itself was just fine. Apparently, only fine when it has partisan benefits.
Either way, the solution here is a tool not available when the VRA was passed, and that's fair ungerrymandered maps. There are tons of algorithms that produce fair maps, certainly much more fair maps vs the current gerrymandered maps. We even have algorithms that can measure the level of gerrymandering.
An anti-gerrymandering bill would be good for everyone and should be supported at the federal level. The only people it's not good for is incumbents.
> It’s unclear to me why transportation demand pricing is allowed but not delivery.
I don't think it should be allowed. It's predatory. It allows a company like Uber and Lyft to see things like "Oh, you are going to a hospital, then I'm going to apply a 10% surcharge because you are probably desperate".
It also works against the drivers. Uber/Lyft see things like "This person is logged on for 8 hours, they are desperate, so let's give them lower rates and worse routes."
Why shouldn’t a company be allowed to price the product differently? For an airline, booking a flight 6 months out, 6 days out, or 6 hours out are different situations.
For Uber/Lyft, booking a ride into the middle of nowhere carries a cost for the driver that isn’t present when booking a ride to the airport.
A flat fee per mile doesn’t make sense. A flat fee per seat doesn’t make sense. Grocery stores already price segment via coupons, sales, and loyalty programs - this is just an extension of that.
> Why shouldn’t a company be allowed to price the product differently? For an airline, booking a flight 6 months out, 6 days out, or 6 hours out are different situations.
Obviously acceptable. But I'm talking about two people placing an order for the same ticket, same airline, same rewards program and getting different prices because the airline knows something about their private life. For example, maybe the airline has gotten information that one person is going to a funeral and thus are more willing to pay a higher price.
Everything about this encourages corporate espionage on private citizens which I'm opposed to. I'd rather not have google calendar secretly sharing my entries with delta so they can in turn determine what ticket price I might be willing to pay.
There's also not market fundamentals which can correct this situation. Back to the airline example, for any given route you are limited by the number of carriers to that location. Airports are physical things with limited capacity which naturally leads to monopolies and oligopolies. That means you can't just "pay someone that's not doing that" as you only have 2 or 3 options for most locations and they all want to participate in that scheme.
You're literally saying "an airline, booking a flight 6 months out, 6 days out, or 6 hours out" is not "charging two people for the same route differently", completely missing the point of alex43578's excellent question.
I'm not. alex43578 was shifting the goalposts from the point cogman10 was making; I acknowledged what he said, but it was besides the point. An airline charging differently depending on the time ahead of flight is sensible. An airline charging differently depending on the buyer's home address is discrimination.
You said "charging two people for the same route differently" is bad: airlines do that constantly and that's why there's dozens of fare changes, fare buckets, sales, codeshares, etc.
Regardless, the bigger point is that businesses already have a ton of levers to move for pricing: sales, loyalty programs, and regular price adjustments. None of those are considered discrimination. Why does the buyer's home address fall into this protected class; particularly for any service that involves transport, delivery, etc to that address? There's a clear relevancy of the address to the cost of a service based around that location.
> sales, loyalty programs, and regular price adjustments. None of those are considered discrimination. Why does the buyer's home address fall into
Because everything you listed applies to everyone equally! Assuming a normal loyalty program anyone can join.
> any service that involves transport, delivery, etc to that address
Shopping at a grocery store doesn't involve that. But sure most forms of charging for transport based on destination are fine. That's different from charging two people differently to go the same place at the same time. "Home address" is just an easy piece of personal info to mention.
(An exception to that most would be like the hospital example, charging more for that specific location inside the general area because the buyer seems desperate.)
I suppose you are misunderstanding me on purpose, but let me try again in very clear terms anyway: Offering the same service or product (a specific flight if you will, a chunk of butter of the same brand in the same store at the same time) to two independent customers at different prices based on prior knowledge about them unrelated to the specific good or service is fundamentally unjust.
What you are referring to is 'price discrimination'[1]. @alex43578 is correct in his examples. In the 'Uber/Lyft' example, his metric for service similarity in the case of a ride to the airport vs. the middle of nowhere can be seen in the distance driven. The problem is that arguments can always be made on why pricing one demographic vs another makes business sense.
In the case of Uber/Lyft, the company can say a ride to the middle of nowhere costs more than a hotspot destination because the odds of finding someone hailing another ride from there are low. This would mean the driver would have to spend more on gas picking up their next customer. Although this seems reasonable, it's probabilistic in nature. This may also not be the case, but the company must price this risk to keep their drivers happy. Well what of the case where the destination is a dangerous neighborhood where the driver feels like their life will be in danger? How do we price the risk then? And that says nothing about the possible mismatch of perception between the seller and the customer.
How about if a grocery store sells goods at a higher price to customers in lower income areas because they notice that it lowers the number of high income area customers to the point they make less profit? Is it right for that store to raise the price for identifiably lower income area customer to make up for the lost profit?
> Offering the same service or product (a specific flight if you will, a chunk of butter of the same brand in the same store at the same time) to two independent customers at different prices based on prior knowledge about them unrelated to the specific good or service is fundamentally unjust
Your statement includes things like loyalty programs and memberships. Presenting these credentials at checkout means customers are willingly giving the company "prior knowledge about them" (that they've shopped at the store before and how much they're willing to spend) unrelated to the *specific* food or service they're purchasing. Should these practices be allowed?
The point of this reply isn't to say what should or shouldn't be allowed, it's to show that I believe the issue is more nuanced than you can account for in your statement of what constitutes unjust business practices.
Why is it unjust? It’s absolutely the store or individual’s choice to charge what they want to who they want, assuming that they aren’t discriminating against a protected class.
In your example, why aren’t all prices then fixed between different stores to ensure justice? Whole Foods shouldn’t be allowed to charge more than Discount Food Bin for the same can of beans, and WF in Oakland shouldn’t charge less than WF in Marin.
> This is something Lyft and Uber do, but it's not really possible in retail.
It is possible for retail. For example, you can simply not display the price. You can display a price range. You can use EInk displays which auto-update based on who's approaching the item.
And of course it's infinitely possible in an online store.
One example of how this is being employed is McDonalds trying to push everyone to use the app. They'll give lower prices in app while raising prices on the menu giving a "not using app" tax. That enables them to have flexible per user prices within the app. A store could do the same thing.
How would that work? The barcode on the item doesn't get rewritten, the checkout counter can't distinguish who picked up which exact item. Even if they did assign unique barcodes to each item, what happens if you take the item off the shelf, and put it in someone else's cart? They'd be charged the wrong price for the item.
The way McDonald's does this that they essentially offer discounts that you can only get by using the mobile app, and the mobile app can dynamically price your cart however they wish. All they (or any other business) has to do is to increase their non-mobile prices to squeeze out non-identifiable shopping.
> the checkout counter can't distinguish who picked up which exact item.
Sure it can, cameras and facial recognition.
Identify a person when they enter the store. Track them with overhead cameras as they move throughout the store. Set the price for the individual when they get to checkout. It barely requires any extra infrastructure in most stores.
You don't need a unique barcode per item, you simply need to be able to identify the person purchasing the item at checkout.
Barcodes today don't contain price information (except for meat AFAIK. But then you'd mostly just need to change it to be weight instead of price).
The store could make it mandatory to swipe your fidelity care before calculating your prices. They already do something similar with specialized promotions.
If we're including promotions or membership discounts, then coupons fit this definition of price discrimination. And those have existed since the 90s at least.
The problems show up when different people don't have access to the same coupons published by the store. Most coupons are fine.
They haven't had the ability to do the significantly bad kinds of targeting until recently. This is a new problem even if it's similar to old practices at the surface level.
In the UK, it's common for supermarkets to have loyalty cards (e.g. Tesco Clubcard) and then they display prices along with the price for using a loyalty card. Then, they just apply the discount (which varies according to the product) when you swipe your card before paying.
Refreshing the content on an electronic shelf label (ESL) takes about 30 seconds, and multiple people can view a product simultaneously. Unless the store is giving everyone AR glasses, people will notice the price discrepancy.
This assumes you have sufficient data to actually recognize a shopper such as facial ID or some form of iBeacon for every single product for which you wish to implement price discrimination. Basic ESLs cost $3 to $12, depending on size and use very little energy. Adding a camera means more energy, so a bigger battery and more cost.
Using in-app discounts is the most likely way to implement this, which I am okay with. Shoppers are willingly trading their data privacy for a discount.
I think viewing it at as a discount is framing it wrong. It’s more a fee for not using the app, and if you use the app you’ll get charged the highest price McDonald’s has decided you will pay.
Should this be legal is a question you could argue both ways, but in my opinion society will be worse off with per customer pricing.
> Refreshing the content on an electronic shelf label (ESL) takes about 30 seconds
Today's, yes, but that's really not a technology limit. EInk displays have much faster refresh rates. The main limitation is the communication, but that's a solvable problem with a smart enough networking mesh.
> multiple people can view a product simultaneously
You track the people's position. If multiple people are crowded around the same product then you can pick a price for both customers. Might not work well for crowded stores, but in that case you probably can just maximize the prices anyways as a lot of people are currently shopping.
> This assumes you have sufficient data to actually recognize a shopper such as facial ID or some form of iBeacon for every single product.
No, you just need 1 customer ID on entrance to the store. Everything else can be done by tracking their position with overhead cameras. You don't need to put a camera on every product.
> Using in-app discounts is the most likely way to implement this, which I am okay with. Shoppers are willingly trading their data privacy for a discount.
I'm not OK with this. Simple reason, it leaves the wide masses with no other option than to sell their data to survive.
> Using in-app discounts is the most likely way to implement this, which I am okay with. Shoppers are willingly trading their data privacy for a discount.
Your mistake is assuming it's a discount, when it's not. For example, Safeway near me charges exorbitant prices for goods which are anywhere from 30-50% lower in the app. What they're doing is the same as your average dark pattern, you're only getting the real price using the app otherwise they charge a no-app fee. And even then you can't tell what the real price is supposed to be, because the app will tailor discounts to your shopping behavior.
Shoppers can and have noticed the price discrepancy [1] which is why this legislation is happening in the first place. If the price isn't the price then the whole basis of capitalism and consumer choice falls apart because there's no way to make a proper determination if Store A is cheaper than Store B.
I think McDonalds dynamic pricing is great. Every time I checkout the app there is some crazy deal. Sure its not always something I want but I'm not necessarily competing w/ the other items on the menu. If there's no deal on something I want, I check BK or similar chains.
> The best example of why data collection isn’t the bottleneck is Tesla. They boast about billions of miles of data, yet they’re struggling to put out fully autonomous vehicles.
Well, TBF, the tesla data was complete garbage with earlier vehicles. They had cheap and somewhat bad cameras in the earlier vehicles that was only somewhat recently updated. And even then, I don't think Tesla is at the end of their hardware journey. I think they don't think that either, which is why they've gone to a subscription only model for self driving vehicles.
Waymo, on the other hand, has gathered less data, but more high quality data. They do the expensive mapping of a city which is a big part of why their vehicles have early on been able to do some pretty impressive feats. The drawback is getting that high quality data takes a lot of time and resources.
> And even then, I don't think Tesla is at the end of their hardware journey.
I dunno about that. Tesla seems completely adrift, pretending to pivot with random forays into humanoid robotics or whatever, to the point that I wouldn't be surprised if they exited the consumer vehicle space altogether within the next decade. They have no answer for Chinese competitors.
I recently watched some videos related to the production of cybercab, which has now started public testing. They’ve still done some great engineering, to the point that the car is now assembled like a matchbox car. All the drive components are contained in a single package for a FWD configuration that the body just drops down on. The car now has no controls besides the screen and door pulls. The materials are all lower cost and they even found a way to skip painting the cars. All of this should help them cut costs significantly.
As far as the self driving, they may be far off still, it’s hard for me to get a read on that and this vehicle is a bet that they will be able to achieve it - right down to the braille in the cabin, so maybe that’s why they still fail. The thing I will say is that despite the PR disaster that the CEO is, which gives us that feeling that the company has lost its mind, it seems they are still quietly doing some advanced engineering.
> Or are you just saying the person placing the cameras is decoupled from the person making the decision to aggregate them all.
That's exactly what's happening.
People are buying webcams which are cheap and have in their ToS something to the effect of "we get to sell everything the camera can see". Which, in turn, allows them to partner with Flock and sell video footage directly to them.
Consider the fact that at one point, Amazon partnered with Flock to sell their ring camera footage to Flock. [1] It only got botched because of the creepy superbowl commercial selling the spying as "finding lost puppies".
At home, we had a beekeeper that'd keep his bees on our land. He'd give us 5 gal drums of honey on request (no joke) which lasted years for us. The honey in those drums would crystalize and harden on top, but with just a bit of stirring it'd re-honeyify. It always just tasted like honey. I never noticed a difference as it aged.
The other thing is that common issues are common. I have to wonder how much that ultimately biases both the doctor and the LLM. If you diagnose someone that comes in with a runny nose and cough as having the flu you will likely be right most of the time.
reply