I’m sorry that you find the modal experience in reading groups to be bad?
It seems like the reading group that the OP put together was really successful. Most are not. That’s not really the fault of someone else for sharing their experience.
Can you give some concrete examples of “cheap young” contract workers being hired to work on product features? You seem to know a lot of things so maybe some concrete examples will help.
In what world is AWS great?
A world where they don’t have a clean way to group resources and clean them up? Their own twisted identity products in AWS that doesn’t integrate outside their cloud or their other products. A world where they leave lingering resources when you cleanup something.
AWS has to much hype riding behind it.
It’s not enterprise class and looks very incohesive.
The point about lack of projects on AWS I agree with if that’s what you are talking a about. That said AWS is absolutely a cloud leader. We use azure and AWS now and there continue to be reasons AWS does well despite cost and quirks
This is certainly a troll post. I use all three cloud providers. GCP is the worst of all, just try their simple text to speech UI. It doesn’t work most of the times.
Don’t even get me started on a deployment story for GCP their deployment manager is deprecated and redirect you to use terraform.
I hate to swallow it but Azure was more usable and straightforward.
Azure has had multiple global outages for many services. I believe you can look at overall stability and see it's not in the same league as GCP and AWS.
Lol can you use ed25519 keys with Azure yet? I doubt it. It was and will continue to be a joke because MS can't pay or promote people that care before they leave.
Azure’s target customer group is non-tech enterprise. The type of group who generally buys solutions over builds them in house. Where needing ed25519 keys is not a common ask
This might be dumbing this down very very much but it all boils down to having some sort of a special lookup table and you give it a "query" and a bunch of "keys" and the "value" is the most likely next word. As you input something into the network, a network of these tables are consulted and you are given most likely next word.
The novelty in this paper is this "query-key-value" relation that gets learned. A lot of previous work in this area was focused on learning a rough state machine to which you input a set of state transitions and it will give you the most likely next state. This will also work but training such networks is very slow and you also don't have the capability to train the network to "attend" to certain part of the inputs. This lookup based technique lets you do that plus this is also very compute efficient (compared to previous techniques).
I'm missing a lot of details but that's basically the intuition behind this.
I call bullshit on this. I had a very rare heart condition that was discovered during one of my yearly check ups. Chasing the root cause led us to even more scarier findings.
How many examples of bad outcomes would it take to retract your "bullshit" assessment?
Ideally, somebody would take all the good that came from the checkups, and compare that to the amount of bad effects where poor interventions were chosen, and weigh them out. Such a study would be quite useful...
I’m pretty sure bad outcomes out number good ones. Were there bad outcomes because of an incompetent provider or a test with high false positives? I’m pretty sure it’s the former.
As always, population-level recommendations don’t result in perfect individual recommendations. Your condition was very rare, and the test is not perfect. Thus if it were done routinely, many people would be flagged as a false positive and treated unnecessarily to catch your one case.
Sure, if we are dropping anecdotes, my kids were over diagnosed/incorrectly diagnosed. It ended up costing me slightly under 1 thousand dollars before other doctors said 'no big deal'.
The weirdest part about these, both of the diagnosis seemed like there was no possible solution, so even with the confirmatory tests, it wasn't like anything was going to change.
However they were insistent of getting multiple specialists on it.
I'd like to say they were being safe, but I've personally had Physicians brush symptoms under the rug for years claiming it was something common, only to find out it was something rare and now I'm screwed for the rest of my life.
Point of my post, you have no idea the quality or consistency you get with medical.
With kids it's problematic. A doctor will see thousands of kids with acute nothing burgers and then one day a kid with emergent type 1 diabetes will come in. Another bad thing is rate conditions are rare and there are a f'ckton of them.
What bothers me is at least in the US we've forced doctors to adopt an MBA driven pop mass manufacturing system. Like they're some schmuck in a chicken factory.
You lost me there in calling Sankrit the mother language. Telugu has a different lineage and Sankrit is alien to Indian Sub-continent. Please stop using HN to subtly push your agenda.
True. But of all South Indian languages, Telugu shows the most Sanskrit influence.
> Sankrit is alien to Indian Sub-continent
This is a bizzare assertion. Classical Sanskrit was developed in India. That is likely true of even archaic forms of Sanskrit and the Prakrit substrate.
Words from Sanskrit can be found not only in several Indian languages but even in some south east asian languages. I speak four languages out of which three have words from Sanskrit. Learn a bit before alleging things.
reply