ranman

They'll cut you a private offer for bedrock tokens but bedrock has a 32k output limit

Which countries? Density matters but raw volume matters too.

Route53 being off on the side but unconnected is still valuable info...

It's a little ironic that the text wrap doesn't work on some of these examples on WebKit.

Eventually he'll be about tree fitty

we built the LLMs unethically, and that they waste far more energy than they return in value

If these are the priors why would I keep reading?

Caylent has been testing this for quite some time and I have to say it's an interesting take. With claude code you can shift between planning and coding modes and this offers a similar approach. The speed is quite good and the results are solid. The spec approach is solid but it takes a learning curve. Some of the tooling and autojumps take a bit to get used to because they differ from the other IDE approaches.

Overall I do believe this has accelerated our development and I'm interested to see where it goes. I don't think it's a direct comparison to claude code or cursor - its a different approach with some overlap.

Humanity fails that question an embarrassingly large number of times.

You get a scale at AWS that is hard to find elsewhere. There are still a huge number of very smart people there. You can learn a lot. I loved my time at AWS.

That said there are a ton of cons. There's an entrenched management class that is disconnected from reality. There are a number of ~L8-L10 folks who don't believe or understand how they're falling behind the cloudflares and other providers. There is a bizarre arrogance in Seattle that masquerades as "willing to be misunderstood for long periods of time". People aren't afraid enough.

What AWS will struggle with over the next few years is verifying the results of the narratives they tell themselves. At some point along their evolution a disconnect between narrative and reality happened and someone needs to bring everything back to a baseline of reality. Leaders tell a story of their success (that I'm sure they themselves believe) and no one follows through to actually verify the results.

This issue of lack of narrative/reality baseline, to me, is a cancer at the heart of AWS and if it can be addressed then I think they can recover and shine. Otherwise they'll fall into the same trap as MSFT back in the 90s/2000s where they think everything is going just fine while the floor falls out from under them.

Someone mentioned that this took almost as much compute to train as the original model.

I don't get this take... what is the scam part? It's a real device that people can buy, and it does what it was purported to do... maybe it's not as amazing as people hoped, but does that make it a scam?

Do 3/4 say it does more good than harm?

I'd love to see one of these for migrating from Claude's prompt API to their messaging API

Be lucky?

I don't have any particular advice. I've been very lucky to join a few companies that were on the run up at the time but I never really knew at the time that they were going to be successful.

I just think blindly pursuing passion to the exclusion of profit will leave most people happy (or not) but impoverished and some people happy, lucky, and wealthy.

It was really just a tongue in cheek comment. I think Paul's advice is inspirational but it should be checked with a healthy dose of reality.

Don't try to guess whether gene editing or LLMs or rockets will turn out to be the most valuable technology to know about. No one can predict that. Just work on whatever interests you the most.

This optimizes for Paul's outcomes not for the individual outcomes.

Is this why AWS had this announcement on Friday? https://aws.amazon.com/blogs/aws/mistral-ai-models-coming-so...

If this clears anti-trust but Amazon/iRobot didn't then we'll know the anti-trust side of things is just based on your lobbying spend.

... why?

Every single one of these could be engineered away.

You clearly didn't read half of the article. 50% of these are purposeful chimes.

Also, why? What is the value of making a compressor or fan more quiet on the 3? It would be much better to lower the cost. Worry about the sound on the higher end vehicles but let the 3 be low cost.

If you want to use this but are tired of waiting for the SDKs to get updated you can do this:

```

curl -sS https://cdn.caylent.com/bedrock_new.zip > bedrock.zip

mkdir -p ~/.aws/models/

unzip bedrock.zip -d ~/.aws/models

```

Now do it with the 2021 icons

A simple, high-throughput file client for mounting an Amazon S3 bucket as a local file system.

SVB represents $200B assets and the customers are in large part revenue negative or growing startups. That doesn’t make it Ok but it cannot be representative of the greater economy.

There are ~1900 publicly traded companies with a market cap over $1B. (NAMER) There are over 1000 privately owned companies with valuations over $1B. (NAMER?)

Writing off startups as not contributing to the economy (jobs, spending with vendors, etc.) is a harder to argue in 2023 than it would have been in the past.

This very well could be the beginning of a stall in the greater economic engine. Except this time it isn't necessarily fueled by failure of the underlying assets. Instead the underlying assets may fail because of the banks. Which could cascade into additional failures across the system.

The parent comment is neither factual nor advisable.

You build greenfield in cloud precisely because it is greenfield and the utilization isn't well understood. Cloud options let you adjust and experiment quickly. Once a workload is well understood it's a good candidate for optimization, including a move to self managed hardware / on prem.

Buying hardware is a great option once you actually understand the utilization of your product. Just make sure you also have competent operators.

This is why we can't have nice things.

The amazon recruiters are some of the most idiotic humans you'll ever meet. At least in NAMER. When I was hiring at AWS I would actively try to avoid the candidate talking to the recruiter too much because the recruiter would frequently say things that were entirely false.

I would judge a vendor or consulting firm based on their marketing website. Why wouldn't I judge a car maker?

From the article: "Police and prosecutors generally don’t expect much evidence from Echo recordings — if any — because Echo speakers are activated with a wake word — usually “Alexa,” the name of the voice assistant."

They complied with a legal order - they didn't submit it to the FBI.

Do you need some more tinfoil for your hat?

Then your mind is incredibly easily boggled and your threat model is broken.

You can't reasonably compare a housing vacancy tax with a hyperbolic straw man argument of a theater. It's not a similar phenomena at all, because it doesn't happen.