Some error corrections: ebay bought skype in 2005 for 2.6B (1 million concurrent user), and Microsoft bought them in 2011 for 8.5B (22 million concurrent users) with 74 million concurrent users reported in 2013 falling to apparently 30 million now...
Microsoft had essentially left it for dead shortly after the acquisition. There were multiple assassination attempts from Lyncs-become-Skype-for-Business-become-Teams department and I can only assume noone to fend for the foster child.
Skype under Microsoft had only seen sidelining and feature & platform regress, while it was a thriving product during eBay ownership. It may not be obvious from the bar graphs but it was apparent when one used the product.
Yes, that would be complicated. EXCEPT most of what amazon sells is Company A in China makes product -> Company B imports and sells on amazon. That's easy to calculate.
A and B passed along the tariff costs to C, which passed those costs to D. D now costs more, assuming they pass the delta cost to the consumer. The only delta a customer cares about is the past/current cost of D, which is what amazon would presumably display.
Isn't this the horn everyone keeps trumpeting? It doesn't seem complicated.
There’s a million reasons besides tariffs a cost may increase. Sellers with no tariffs will just increase prices to pad profits if they’re given an opportunity to let Amazon pin it on tariffs. There is a lot of evidence supporting the theory that this type of manufactured inflation is what drove the recent pandemic era inflation. Corporate profits alone are the best indicator of this occurring. They raised prices by a lot more than their cost increased because they could hand wave broadly at inflation.
A lot of things go counter to economic theory. If you’re an American company used to selling 30% above Chinese competitors, you likely can’t ramp up production quickly or will hesitate to make the decision but will likely increase your prices simply because you can and partly because you’re used to the charging x% over Chinese goods. You may even feel being price competitive with China lessens your perceived value, quality, etc. in the consumer’s minds.
It is complicated because tariffs are not just "passed off". In the real world, tariffs are almost always marginal, meaning each company is willing to absorb some of the cost, ie lower profits. In parent's example, the US has a large steel industry and rail companies relying on imports cannot arbitrarily raise prices because they need to stay competitive with companies using domestic. Markets are also dynamic and simply comparing pre/post price may work temporarily, but for a persistent tariff that metric will become increasingly meaningless as price inevitably fluctuates due to other factors.
Prices fluctuate all the time especially on Amazon items. Inflation is also a factor. Just taking prices from Jan 19, 2025 as a baseline and attributing all other changes to tariffs isn't going to be accurate.
This doesn't help; e.g. if you buy sneakers made in the US, those might be affected by leather import tariffs, but Amazon is not gonna know by how much and manufacturers are gonna be cagey with exact numbers.
Remember these are elons are script kiddie hackers, it only occurred to disable the outer firewall, azure ad will independently geoip block all by itself
It buys you approximately two days (with reservation discount) of a single p5.48xlarge instance, which has 2TB of RAM, and 640GB of VRAM in 8x H100 cards. In fact that is the pricing example they use: https://aws.amazon.com/ec2/capacityblocks/pricing/
MI300X (RunPod) 192gb ram
Hourly Rate: $2.49/hr.
Break-even Point: You can rent for 2,410 hours (~100 days of non-stop-continuous use) before reaching the cost of the $6000 Mac. Mac's top out at 192GB not 2TB ;)
Consideration: If your AI training requires sporadic use (e.g., a few hours daily or weekly), renting is significantly cheaper.
MI300X will also get you result many times faster too, so you could probably multiply that 100 days!
So based on the 4090 prices and the small footprint, great thermals of the 5090. It doesn't seem like a bad deal to get 32GB VRAM and the fastest inferencing performance for $2K. Its looking like a 40% increase in performance (not that the 4090 was slow by any means).
https://www.statista.com/chart/1417/skype-usage/
What is your criteria for stewardship ?