Perhaps, but sending everything to the cloud might get them in (very expensive) trouble. Depending on who we are talking about, of course.

cost isn't even close to the main motivating factor for my context

Or have an a11y standard for MCPs, where they can't show UI elements and have to only respond with text so that Voice Readers could work out of the box.

This would be a game changer, currenly Voice Readers do not work very well with websites and a11y is a clunky set of tags that you provide to elements and users need to move around elements with back/tab and try to make a mental model of what the website looks like. With MCP and Voice chat, it is like talking to a person.

Agreed that better browser support for accessibility tools is always welcome. I just don't think MCP is required at all there, the APIs are already well documented and built right into the browser.

In a real world test you would have a tool call for the LLM which is a bit high level like GoTo(object) and the tool calls another program which identities the objects in frame and uses standard programs to go to that.

Replicating features from existing software has become extremely easy due to AI. I won’t be surprised if open source is able to easily catch up with the bigger products.

Replicating the software is easy, running the services at AWS-scale is hard.

Would you need to though?

If an organisation ran it's own instance, it would only need to scale for that organisation ( including any external attends over a bridge ).

That does of course assume companies have the expertise/appetite to run things themselves.

lol it's the flask debug server, "don't use this in production" banner and all

It is absolutely the worst time to be a gamer. First it was the GPU prices that went up and NVIDIA started to focus on their enterprise cards more and more RAM prices. I don’t think I’ve seen the price of computer components go up so much.

Imagine if historic humans had decided that only hammers are enough. That there is no need for a specialized tool like Scissors, Chisel, Axe, Wrench, Shovel , Sickle and that a hammer and fingers are enough.

Use the tool which is appropriate for the job, it is trivial to write code to use them with LLMs these days and these software are mature enough to rarely cause problems and tools built for a purpose will always be more performant.

It is absolutely crazy how much AWS charges for data. Internet access in general has become much cheaper and Hetzner gives unlimited AWS. I don't recall AWS ever decreasing prices for outbound data transfer

I think there's two reasons: one, it makes them gobs of money. Two, it discourages customers from building architectures which integrate non-AWS services, because you have to pay the data transfer tax. This locks everyone in.

And yes, AWS' rates are highway robbery. If you assume $1500/mo for a 10 Gbps port from a transit provider, you're looking at $0.0005/GB with a saturated link. At a 25% utilization factor, still only $0.002/GB. AWS is almost 50 times that. And I guarantee AWS gets a far better rate for transit than list price, so their profit margin must be through the roof.

> I think there's two reasons: one, it makes them gobs of money. Two, it discourages customers from building architectures which integrate non-AWS services, because you have to pay the data transfer tax. This locks everyone in.

Which makes sense, but even their rates for traffic between AWS regions are still exorbitant. $0.10/GB for transfer to the rest of the Internet somewhat discourages integration of non-Amazon services (though you can still easily integrate with any service where most of your bandwidth is inbound to AWS), but their rates for bandwidth between regions are still in the $0.01-0.02/GB range, which discourages replication and cross-region services.

If their inter-region bandwidth pricing was substantially lower, it'd be much easier to build replicated, highly available services atop AWS. As it is, the current pricing encourages keeping everything within a region, which works for some kinds of services but not others.

Even their transfer rates between AZs _in the same region_ are expensive, given they presumably own the fiber?

This aligns with their “you should be in multiple AZs” sales strategy, because self-hosted and third-party services can’t replicate data between AZs without expensive bandwidth costs, while their own managed services (ElastiCache, RDS, etc) can offer replication between zones for free.

Hetzner is "unlimited fair use" for 1Gbps dedicated servers, which means their average cost is low enough to not be worth metering, but if you saturate your 1Gbps for a month they will force you to move to metered. Also 10Gbps is always metered. Metered traffic is about $1.50 per TB outbound - 60 times cheaper than AWS - and completely free within one of their networks, including between different European DCs.

In general it seems like Europe has the most internet of anywhere - other places generally pay to connect to Europe, Europe doesn't pay to connect to them.

I’m more curious to understand how we ended up creating a single point of failure across the whole internet.