angstrom

Well, for problems that are simply looking for a ranked relationship if you have human input it could be used to train a ML that attempts to look for similar correlations...or you could just use your human mechanical turks that are already informing you. ML problems that are good are the ones not trying to approximate reality like CWBAB.

I'm doubtful we'll see an ai that makes a serious jump without directly interacting with the world we live in. Under that measure cars might be the closest since you learn to interact with the bounds of where it can and can't go is similar to a toddler learning to crawl.

Dirty secret: "The customers who bought also bought" algorithm doesn't require ML/AI.

You can accomplish that with relational algebra on a precomputed data warehouse job and only for products with strong correlation. The intelligence of the customers is enough agency to instil a semblance of intelligence in the data.

I'm in the NYC market. I use Lyft to get around and Uber Eats to order things that could never have a business model delivering to me. To me market capture is just as bad as the original Taxi model, but even Taxis have gotten around to handling this problem. You can switch to the app in the middle of a ride and pay with the code on the back seat tv screen. The convenience factor is only there for food until another competitor enters that arena.

It's a race to the bottom and I suspect after all the hard work is said and done Uber will be the Yahoo of ride share. Tesla may very well swoop in and say "Thanks for blazing the trail"

Amazon sees a lot more and all they really care about is what you view and what you buy. Everything else is noise with current tech.

*Shortage of qualified people.

I pity the people that have to filter that.

This has always been my interpretation. I was unaware the idea was to sell C02 heavy companies.

I've seen coworkers die in their 30s, 40s and 50s from cancers and heart attacks. Can concur.

It's easy to ignore. Just as easy as 2003 when the American empire secured the ME oil unilaterally from Iraq by ousting Saddam while all other powers made no retaliation for the move other than strong condemnation which is really the same thing. No one liked it, but no one was really going to lift a finger to prevent it.

https://en.wikipedia.org/wiki/Opposition_to_the_Iraq_War#Off...

That's not the same. This is more along the lines of examining what is causing the negative CP. The margin is impacted by several factors that determine the cost, but then you can go a step further and look at the net impact on profit based on the items it was purchased with. Much of what is sold on Amazon is not stocked by Amazon. Therefore the symbiotic relationship that is best is to maximize vendor profit AND Amazon profit. This means being competitive on the Amazon marketplace as well as against other retailers and that can be best achieved by Amazon lowering the net margins to be profitable, but not uncompetitive. These are things from the quality of images and detail about the product to the shipping options, to the damage incurred during storage/shipping. The end goal is to optimize for the customer propensity to checkout and not return the items which is also a good sign of customer satisfaction. Most people who walk into a grocery are going to walk out with whatever they put in their cart. Cart abandonment is much higher in e-commerce and so there's a good deal of effort understanding what lowers the propensity to complete the order.

I’ve worked for Amazon. The thing you have to remember on the retail side is much of the focus is on CP (contribution profit) and much of analysis has to do with finding the largest negative CP products in different marketplaces and pulling levers to make them less negative. There’s a large number of negative CP items carried just because it improves customer selection which helps drive larger cart orders. I never heard of squeezing more profits out of already profitable products probably because positive CP is an indicator of a competitive price point.

It's also worth remembering that innovations in manufacturing are quite often 'short cuts' that have the same or better result with less process.

Fair point. It depends on what the project is.

After 60 years of advancements in networked computing we should see not just the performance benefits, but also the lifestyle benefits. I don't see the point of sitting in an open plan office especially when you have cliches of people that choose to speak in native languages about technical subject matter. I'm not benefitting from the distraction, nor is anyone else.

I would probably go with US-EAST-2 just because it's isolated from anything except perhaps a freak Tornado and better situated on the eastern US. Latency to/from there should be near optimal for most eastern US/Canada population.

Netflix actually added the additional AZs because of a prior outage that did take them down.

"After a 2012 storm-related power outage at Amazon during which Netflix suffered through three hours of downtime, a Netflix engineer noted that the company had begun to work with Amazon to eliminate “single points of failure that cause region-wide outages.” They understood it was the company’s responsibility to ensure Netflix was available to entertain their customers no matter what. It would not suffice to blame their cloud provider when someone could not relax and watch a movie at the end of a long day."

https://www.networkworld.com/article/3178076/why-netflix-did...

It’s refracted light arcing in the atmosphere. That’s not a hallucination. Treating it as some continuous entity is a hallucination of our minds. There’s nothing suggesting the start of the arc has anything to do with the end, but we call the collective refraction phenomena we witness a rainbow. UFOs could also be something that we collectively fail to comprehend well enough to categorize the constituent parts.

It appears to have an observer quality to it. The one that sticks out the most in my recent memory is the 2006 Chicago O'Hare sighting. I remember this going through a pretty quick 1 or 2 day news cycle. Fairly alarming given the number of people witnessing the phenomenon and the backgrounds of the people observing it.

https://en.wikipedia.org/wiki/2006_O%27Hare_International_Ai...

Obviously rainbows are an example of a shared hallucination and the phenomena manifests differently dependent on the observer's viewing angle. At the moment I'm not convinced these aren't unexplained atmospheric phenomena because I've never seen one. I know some ideas put forth include ball lightning for at least some UFOs. Assuming these cases to involve intelligence may just be our misinterpretation of the phenomena.

When you consider the effort to get to that stage of locomotion it's amazing that flying birds haven't advanced faster technologically.

As much as a I like the movie it's worth noting that on such short timescales there's probably a more significant impact of nutrition differences and the failure of IQ measures across different society structures ie agrarian, industrialized, service oriented.

The long term capital gains loss is funny. I think I ended up with something like 30 years worth of LTCG losses I'll be able to carry forward. It's like a reminder that just keeps on reminding year after year how badly you can get shafted.

They are, but there are still some startups that use options. At least the liquidation is easier with the RSUs, but it doesn't prevent the fall off post IPO as people wait (rightfully so) to see quarterly results employee cashouts before taking a stake in whatever the narrative is.

Switching to RSUs was a good decision, particularly after so many Google employees were borked with underwater options around 2008 that took until about 2009 to correct.

Seems to be normal form. I don't have enough fingers to count the number of times companies go public, primary investors exit, stock slumps. Employees who have been with the company several years get stuck waiting 6 months to sell any shares at which point the stock is well under water preventing an exodus of talent. It's completely by design.

It would be more concerning if these stocks were spiking after IPO because it means the underwriters messed up and the market is irrational.

And a police force that likewise rivals militaries both in size and equipment.

If only it was so clean cut. Tesla isn't even operating in all 50 states due to specific laws in 1/2 which prohibit direct sales. This is part of the reason for switching to service centers and transitioning sales completely to app based purchasing.

Tesla is driving the biggest wedge of change through the world economy in the 21st century. The impact these changes have will affect everyone who uses, sells, services and advertises automobiles. It's no wonder the amount of outright vitriol the press propagates given the lack of advertising Tesla does.

Right. It means exactly what it means. Actual test did not match expectation. Thus, there is an anomaly. Sometimes that means "boom" and sometimes that means bad sensor and sometimes it means the result was outside the margin and they need to do a thorough analysis before retest.

Or the more unnerving and less useful personality trait of lying. The Turing test is after all a game of deceit. Whether the human believes the AI to be human or the AI believes itself to be categorically human begins to blur the lines.

Gateways filter that before internal services see anything. There's not any reason to be playing with passwords beneath the Authentication layer. Credentials should be exchanged for the customer identity and expirable nonce. If that mistake happened in the authentication/authorization layer then it becomes a big question of competence.

That would ignore the potential it represents. Miniscule today. Amazon already dogfoods this product to handle verification of employee badges so you know the person swiping in is actually the person the badge belongs to. It's a real-time two factor authentication. Security is the foot in the door. On the surface that is not a terrible application.

Black box high five for the win, amirite?! Roll it back up to the top of the hill and lets see if it records something different...

It's sad, but most faults seem to be like this. First one is treated as an anomaly, second one is treated as the start of a trend. It happens so many times I'm glad I'm not working with human lives.

As a NYC resident for 11 years now I see this as a necessary transition. The streets aren't getting any wider unless we do away with street parking. I don't hear anyone rushing to suggest that option, although it's easy to forget there was a time when the streets weren't lined with cars - but horse shit. So we trade one inconvenience for a slightly less inconvenient sight (I own and park my car on the street dealing with the inconvenience of alternate side parking twice weekly). The focus really needs to be on improving the efficiency of the MTA, reducing distance between trains with more efficient switching is big part of that. The inefficiencies of the trains and improved efficiencies of app based rideshare dispatch is what is pulling ridership from the trains the most.