dgemm

If AI comes for the trades it seems like it will not be via robots but instead by leveling up all of the DIYers and lower skill handyman types allowing them to troubleshoot/build/fix more than they were capable of before.

They were free in TTC stations.

I find it hard to imagine nvidia's moat not drying up - the hyperscalers already have more cost effective silicon and the AI labs already use a mixture of all the capacity they can get their hands on.

The strongest moat I see is ownership of training data - an area where Google seems to have a clear edge

They bill you per output token.

Never heard of work louder, but it sounds like an idea I used to joke with coworkers about, around making a clickly keyboard with an amplifier and speaker to passive-aggressively demonstrate how annoying the clicky keyboards are in a high density office environment.

They won't necessarily map out all the circuits but they will generally test them all with a tester to find wiring problems.

Becoming increasingly convinced that the pattern that will stick is not AI "integrated" into everything but personal agents. As in I will have one personal assistant that helps me in all of my tasks instead of using the little ones bolted on to every product.

Similar - had an HVAC tech out to diagnose mine (some intermittent electrical problem was killing thermostats randomly) and since it was intermittent they couldn't figure it out. I ended up using Gemini to narrow down a list of potential problem components and just replacing them all which fixed the issue.

Kind of a superpower to turn anyone with a bit of tech inclination and problem solving skills into an HVAC tech - not a very good one, but one with enough motivation to get the results you need

Kinda tired of being inundated with low quality AI slop absolutely everywhere.

Same - 13" macbook screen becomes less functional when you fill it with padding.

State of the art 7 months ago is good enough for a lot of use cases.

I think a lot of usage will move to the cheap open weight Chinese models once there is an incentive to do that. Everything below the highest end frontier models are becoming commoditized and I suspect the commodity segment will pass the "good enough" bar for most applications.

This seems true for our moment in time but looking forward I'm not sure how much it will stay that way. The LLMs will inevitably need to find a sustainable business model so I can very much see them becoming enshittified similar to google eventually making 2) and 3) more similar to each other.

It's a forcing function that ensures the middle layers of a vertically integrated stack remain market competitive and don't stagnate because they are the default/only option

This is the right take - there is a huge variation in "value per dollar" across AWS services. The base ones that solve hard problems like durable persistent state can be very much worth it. They tend to be the older ones.

> AI coding will make us all

I'm willing to believe it will make high-judgement autonomous people more productive, I'm less sure it will scale to everyone. The author is one of the senior-most technical staff at AWS.

I just wish the Gemini app would stop inserting and auto playing a YouTube video into nearly every response when I'm on a mobile connection. There appears to be no way to stop it.

RouterOS 7 with the wifiwave2 package supposedly improves on this by (finally) supporting 802.11r/k/v for roaming between APs.

I don't have any mikrotik hardware new enough to support it so I haven't tried it myself yet and documentation is (as usual) pretty lacking, but like you I want to believe.

It's related to vehicle weight, electric cars can be significantly heavier.

It's super useful for maintenance, for example you can replace and upgrade the drives in place without reinstalling the system.

EPA and DHA actually originate in algae and seaweed which is where the fish get it from, so those are good sources too.

Most of these are what I'd call "cloud in name only" providers - everyone uses the term but you would have significant challenge moving a cloud workload that makes use of the higher layer abstractions to these.

I think the point was not that we won't still use a lot of hardware, it's that it won't necessarily always be Nvidia. Nvidia got lucky when both crypto and AI arrived because it had the best available ready-made thing to do the job, but it's not like it's the best possible thing. Crypto eventually got its ASICs that made GPUs uncompetitive after all.

I think less of that and more of real risks - Nvidia legitimately has the earnings right now. The question is how sustainable that is, when most of it is coming from 5 or so customers that are both motivated and capable of taking back those 90% margins for themselves

The M1 macbook did feel like a big jump for me because it had great performance and no fan, which was a huge thing compared to everything else at the time.

You would be the best one to evaluate if this applies in your case but in many cases where my users say "it's not possible" I end up finding a gap that's more related to usability than technical. I often still find there's something worth learning from this kind of feedback even if it's "wrong".

I'd be curious if you have any insight on why the tracking apps for buses are so inaccurate - OneBusAway but also Google maps and the others. The estimates are often off in one way or another and sometimes a bus just doesn't show up at all. I assume the problem is with the source data but I'd be curious if you have insight into just what causes it to be so unreliable. I'm in Seattle FWIW

The theoretical best latency would be something like speed_of_light_in_fiber/great_circle_distance_between_regions, both of which are pretty easy to find. The first is a constant you can look up, and the second you can compute from coordinates of each region pair.