marcinzm

Fun thing about a two sided marketplace. It needs two sides. If you make life hard for sellers then they won't join and then you, by definition, don't have a marketplace.

> Re: fulfillment — why not ask Etsy sellers to pre-stock everything that's not made-to-order into fulfillment centers like Amazon's? And, in fact, to send anything that is made-to-order to a fulfillment center first, where it'll be inspected, re-wrapped and re-shipped by the fulfillment center — with the buyer's card only having a hold, not a charge, on it until the seller's goods are inspected by the fulfillment center as good? (In other words: turn the process into a goods-for-money mediated escrow.)

Because that costs money and most Etsy sellers sell very few things. We're not talking thousands or hundreds or even dozens per months. We're talking single digits per month. Per store. Not per item. With 7 million stores.

> Re: quality / drop-shipping — I feel like a lot of this could by solved just through simple semi-automated verification. Require for a product to be listed, for the seller to upload a video of themselves making the thing. Then pass those videos to MTurk workers to evaluate them for tricks.

See the previous point. No one would use Etsy if they had to make a whole video and go through an automated process for an item that might sell once per year.

edit: Amazon sells something like 350m products including the marketplace and 12 million directly. Etsy sells over 100m. Amazon has 40x the sales revenue. The economists are so different between them it's not even the same universe.

> The automatic enrolment in the Etsy advertising program means the platform fees for anyone selling more than 10k USD increase sharply at that point to ~20% of total revenue.

That seems to assume a very very unrealistic 100% of orders being from offsite ads. In reality it's probably 10% for most shops if even that many.

> and nothing to stop them

Except Ebay and Amazon already existing.

I find Google worse every day. The latest thing I noticed is a ton of AI generated sites at the top that "summarize" things. Mostly the same format. Except they're usually wrong in many ways. One site kept saying Llama 3 is a Mixture of Experts model despite it not being one.

The goal of the CEO is to drive the stock price up in whatever horizon investors care about. Usually it's not a long horizon. It's not to build a sustainable product or even a product at all.

No but lots of other people have: https://tmsearch.uspto.gov/search/search-results

Feel free to release a computer named Apple to prove me wrong.

That doesn't mean in any way that it can't be a legal trademark.

Moonshot to me indicates there is a specific goal versus simply making a startup that may pivot utterly. If the Apollo project ended up building a high speed train then it'd have been a failure no matter how good that train was.

That's how you run a large corporation. Not a successful startup or moonshot project.

It is perfectly valid to ignore a risk and then assume you'll figure something out if it comes to pass. Few things happen quickly so you'll have time to talk to other founders and investors and so on. Or maybe you'll fail and then start a new venture in six month. Trying to plan for mitigations ahead of time will mean you spend all your time on mitigating risk versus growing the business. A climber that does that is a corpse.

Something can be good for X overall but be bad for subsets of X. The fact that you find it so offensive for me to even make that statement says more about the mono culture you view the world through than the one I view the world through.

edit: Of course Google, the king of the tech bubble, saw this as a positive so your statement about this being a tech bubble thing is also really amusing. If anything the tech bubble would likely agree with you so you're part of the tech bubble mindset on this one.

The mountain climbing is also amusing since, in my experience, running moonshots needs someone who is inherently lazy. Driven, yes, but also lazy. You don't want someone who climbs mountains as a hobby. You want someone who figures out how to get a helicopter to take them to the top of the mountain. You need to de-risk every single thing possible, even if others say you can't do that, by taking the laziest and least risky approach. Then concentrate on the areas you can't de-risk. Then once you get it running you iterate and optimize the lazy bits.

edit: Actually de-risk and not simply lie about de-risking. Eventually you need to publicly test the prototype and the lies will fall apart.

> McKinsey & Co.-trained entrepreneur who abandoned a Ph.D. in theoretical physics to build startups in sub-Saharan Africa

That does not sound like the profile I'd want to run a moonshot project. To me you either you want someone who is very goal oriented to do a research lab approach or very product oriented to do a deep tech startup approach. Abandoning a PhD indicate it's not the former and McKinsey indicates it's not the latter.

Assuming that’s the only clause that they can use to cause people trouble. The article indicates it’s not.

That's an odd straw man. I'd appreciate you not put words in my mouth. You're claiming that self-driving cars are better than public transit and will replace public transit in Europe (assuming they were available). I'm merely saying they won't because that is an inferior solution. I said nothing about manual cars or the shift of manual to self-driving for the segment that does drive cars.

So basically make the urban environment a sea of constant fast driving cars filling all available space. There's a reason a lot of cities are converting roads and urban centers into non-car area. Parking and car lanes replaced with bike paths and green areas. It's a much better quality of life. Which, oddly, some people prefer to a 1800s style corporate dystopia where most people's miserable lives are nothing but a means to drive ever more wealth to an elite minority.

> Self-driving cars will make things way more efficient.

Road capacity has a physical limit. Unlike data you cannot convert cars into light of various wavelengths to pack more into the same physical space. Well you can but that's called public transit. In other words, public transportation has a higher capacity per road/rail area than cars ever will. We don't live in the metaverse after all.

As a result more people using cars (self-driving or not) means you hit the road capacity much more often. That means more congestion and longer trips. Any efficiency you gain by going point-to-point is utterly lost in the increased congestion and traffic.

As the other person noted, European cities are much denser and have significantly less road area than US cities. Thus more cars means significantly less efficiency as you spend more time in traffic.

How exactly will it reduce congestion? The cars are the same size. The trips are the same and are likely to go up in frequency with self-driving. The roads are the same as before. If anything you will have a ton more congestion. The vast majority of congestion is due to not enough capacity and too much demand. Right now that is caped due to the time you waste driving but if it's self-driving then the sky is the limit.

Public transportation increases the potential commuter density per area of road/rail while self-driving doesn't.

The US has 280m with half the population.

Europeans have already solved this problem without need for non-existent technology. It's called public transportation. That's why Americans spend 101 minutes per day driving on average while Germans spend 36.

Imagine how much more GDP the US would have had if they had done that...

edit: Of course, Europeans also use that extra time to enjoy life versus driving GDP that due to increasing wealth inequality won't benefit them in any tangible way.

Did you read what OP wrote? It's not mostly about wearing your seatbelt when seated but mostly about never ever ever leaving your seat. That's the constant fear part.

> - These conditions are unpredictable and have little warning

Not true at all. The vast vast majority of turbulence incidents are very well predictable. It's not 1930 anymore.

> - The conditions that lead to uncomfortable or dangerous turbulence are not at all uncommon.

Except your flawed logic in the previous point flows to this point. The question isn't turbulence but unpredictable severe turbulence which is much rarer. About 1 per million flights I think which is far from common.

> A smart, independently minded person, would do the following risk analysis:

Given the flaws in the risks analysis I find this sentence hilarious.

Because there's millions of those tiny acts and I doubt you've read the risk analysis of each one to decide which fifty thousand to follow. But if you actually do consider them rationally and act on them then cool. Most people who say what you do however don't but rather act based on an irrational perception of risks usually due to things like media attention or misclassifying lack of control as risk.

I've met people who actually and rationally consider all the potential risks and I respect them. It's also an exhausting life to lead from my observation.

That seems odd since I also don't see how this differs from other approaches being published. Except what everyone else calls an Image Encoder (ie: some type of pre-trained VAE architecture) they call a tokenizer. The Apple MM1 paper used ViT-L for example for it's image encoder and then C-Abstractor for it's image tokenizer.

But then it's not about safety. It's about your need to feel in control of the situation even if the control is an illusion. That is fair in a way but saying it's a safety thing or that people are crazy for not needing to feel in control all the time isn't fair. Or even calling them crazy for not doing the same ultimately meaningless actions as you to create an illusion of control for themselves

When I walk in a city I don't spend my time constantly scanning the sky and every building around me for a potential falling object. Or research the structural history of every bridge I drive over then spend hours avoiding certain ones. Or a hundred other rare things that do kill people but statistically not often. Everyone dies one day and everyone has a single life. I don't see the point of spending that one life in constant fear of the end that cannot be avoided anyways without even meaningfully changing the timeline of that end.

I would say using an engine around Earth that emits radioactive exhaust that travels at less than Earth escape velocity is the real lunacy.

As should have the IoT people to not conflict with the decades old LORA name used for Level of Repair Analysis.

And LORA stood for Level of Repair Analysis since the 70s.

> Long said fewer than eight percent of the country's 8000 beekeepers were currently affected - and there were 2900 reports of AFB made in the past year to May, down 15 percent on the 3449 reports the year before.

I wonder if that's just due to people not self-reporting due to the cost of doing so.

> "Beekeepers have to notify and then destroy hives within seven days after finding the disease - and most beekeepers, including Steve, usually comply with that regulation.

And yup. "Most." "Usually."