...isn't that basically Costco.com? the trade-off is that you can't sell a million different things if you want to ensure quality among all the things you sell
We (us, the people reading here, the demographic building the software and technology and automation) would do well as a community to spend more time on introspection about what ends it serves that tech, which was sometimes promoted as a great equalizer, gets built so often in practice only to make the walls of an aspiring monopolist's fortress more steep with no benefit to anybody else.
Technological advancement by itself merely creates the potential for equalization. Whether that potential is actually realized depends on the culture of society, though. You can use the same basic tech to build Star Trek communism or a cyberpunk dystopia.
We as an industry are seen as defined by crap/enshitification now. The world gave us good faith, cheered us on, mostly excitedly hyped us. And we picked, by small choice after small choice, this.
its silly. itd be like introducing first year programming students to advanced maps/filters/anonymous function syntax, instead of the easier to understand for loops and if/else statements. math's "no true scottsman" approach to teaching only hurts itself in the long run.
I'm not sure if it would be easier to explain a map / filter to a first year student vs implementing the patten manually using a for loop and if statement...
Seems like a pretty easy example to make practically, for map have a collection of things, say balls or black. Pick up each one and do a thing to them, paint them blue for example.
For filter do the same except have two different colour balls, if they are yellow they get thrown away, of they are blue they get put in a bucket.
A for loop doing exactly the same you would need to explain the topic at hand, as well as explain iterating an index etc...
Explaining loops is independent of the concepts of collections though. It's also more general, since map/filter/reduce use some kind of loops under the hood anyway, the fact that probably shouldn't be ignored in education process. Unless of course you go with pure functional recursive iterator, but good luck explaining that one.
Maps and filters also require understanding of higher order functions and the very idea of passing function around as a value. I would argue that implementing map/filter with a loop and then demonstrating how this pattern is generalized as .map()/.filter() functions is better and more accessible
Yes! It's a proxy that might modify results on the way in or out, which proxies can do.
Could also be called a gateway, which feels a bit more accurate.
The same way API gateways perform additional services like rate-limiting and authentication and billing, an MCP gateway abstracts the services behind it and adds context such that an LLM can more easily interact with them.
"server" implies that the content being served has the same owner/same scope of control and trust. the sysadmin of an ftp server is the one owning the disk that the ftp server uses; github.com controls the repos that are available on the github site.
i think this whole "mcp security is terribad" thing spawns from the incorrect categorization of the thing as a "server" - if it were instead called a proxy, the rabble would die down.
In relation to the client (AI Agent), the MCP server is serving resources like tools, but in relation to your platform that hosts the API those tools call, it is a client.
heh. you also have to deal with mail theft when you have your own hardware. having a project delayed for 6 months because theives stole a shipment of solid state drives is something i don't want to go through again, lol
the -> cloud -> datacenter -> cloud -> datacenter move will by cyclical every few years.
companies that were in the cloud in 2018, did not have to pay the high costs of fixing hardware that was vulnerable to spectre/meltdown. in some cases, migrating from a datacenter to the cloud was cheaper than upgrading datacenters full of vulnerable hardware.
now with everyone in the cloud, datacenters have lower prices - so its going to draw companies into datacenters. in a few years, though, theyll have to either upgrade their hardware, or move back to the cloud. its gonna cycle like this for the forseeable future.
> Meaning there was exactly zero value in the program.
if your customers die, they stop being customers. definitely some long term value there, but youre right, the psychological damage to the drivers could offset those gains.
I think the point that the parent commenter is making is that if the drivers are always arriving after the actual emergency workers, they're not providing any value. It sounds like they've yet to save a customer with this experiment, so the value you're talking about is still hypothetical.
Amazon is the one performing the experiment; if the sample size isn't big enough to be useful, that's a mark against their experimental method, not the feedback of the people on this forum. The claim that this is actually valuable in some way should be the one with the burden of proof, and it's unreasonable to argue that flaws in an experiment showing one result somehow imply that the opposite is more likely because of it.
reply