Gh0stRAT

>the customer gets more value out of the modded device, which means more sales to the device vendor.

Not sure about the modern xbox, but a lot of game consoles have traditionally been sold at a loss with the expectation of making it up on game licensing fees.

Making more sales at a loss without a significant enough increase in licensing revenue equals a bigger loss. And that's before accounting for the higher support costs of people with incompatible 3rd-party hardware calling Xbox support.

>This does not mean anything more than that the AI has a greater breadth of training background, which is likely.

This has nothing to do with Shannon's definition of information. The light from a star going supernova or a tsunami hitting shore convey information without any kind of agency nor intent being involved.

>What we will not be getting from the AIs is any creative output based on unique understanding, as we would from an intelligent, creative human.

This is a really big assumption here that I don't think is justified. Are you assuming that humans have some non-physical soul which makes us somehow different from any other deterministic information-processing system?

> In classic Information Theory, information is basically surprise value — how much unexpected info is in the message? — yet these "AI" systems put out the most expected subset in each instance.

Forgive me, but isn't this kind of moving-the-goalposts? Information is the surprise value from the recipient's point of view, which meas the recipient's bayesian prior probability is "expected". Saying "these "AI" systems put out the most expected subset in each instance" assumes that the recipient's priors exactly equal those of the model which would only be the case when the model is talking to itself. (or I suppose to an even more complex model with perfect knowledge of ChatGPT's weights)

The fact that no information is transferred when the model talks to itself should not be surprising and would apply to any AI. (even including a superhuman post-singularity god-like AI)

These days it seems like management tries to keep a buffer so they can operate for a little while during a government shutdown. That being said, there are a variety of funding sources and it wouldn't be fair for some teams to be allowed to keep working while others could not, so they'll typically shut basically the whole lab down once they run out of extraordinary budget measures.

Also, if there is an actual shutdown Congress usually retroactively pays the federal employees who were furloughed but it doesn't typically apply to contractors. (DOE labs are staffed by contractors rather than government employees so that they can pay more than the government pay scales)

>It can replace queries like "what's the tallest mountain in France", but that's not really a profitable query for Google.

Being able to answer unprofitable questions has value insofar as you become peoples' go-to search engine for more profitable queries. That being said, I agree the threat to Google is probably over-hyped. (simultaneously, the spectre of LLMs coming along and giving people a reason to visit a competitor is probably the biggest threat to Google in the past decade+)

It works both ways though:

>ChatGPT, summarize the following bill and be sure to note any unexpected clauses: <text of bill here>

>Second, I think the "money well spent" part is that you helped her in some small amount, when it was obvious she was suffering, and you didn't expect anything in return.

Let's be clear here: the "out of gas" story was almost certainly a lie and OP knew it. He helped her buy a small amount of drugs.

I have a sibling struggling with drug addiction for the past 6+ years, constantly "borrowing" money from others and never quite hitting a serious enough "rock bottom" to decide to get help/rehab. My brother "borrows" $10+k/year from my dad and it's poisoning their relationship and draining my dad's finances. (my dad recently took out a 401k loan because he's so cashflow-negative and that money's gone already) My dad enables him every day while dreaming about moving to another state and cutting off all contact.

$40 for OP to cut ties with someone like that is indeed money well spent, IMO.

It depends how many requests (both frontend and backend) each user triggers and the odds of a big latency spike on at least one of them.

For single requests, targeting 95% would probably be just fine back in the days of static sites served by a single Apache instance, but that doesn't describe many modern systems. Nowadays, a single HTTP request to a load balancer can trigger a flurry of blocking queries to fulfil it. (eg a distributed ElasticSearch query that can't combine the results from each node and return until EVERY node has responded) then your worst-case performance will quickly begin to dominate.

I mean, if we're talking about what might've been then there are easily a hundred million lives lost each year to menstrual cycles in the US alone. Just gone. Think of how many Einsteins, Mozarts, and Davincis have already been lost to us! Abortion is <1% of this much bigger problem, practically a rounding error.

^This may sound ridiculous (I certainly think it does) but I think contrasting it with your post is a good thought experiment. Can you see any meaningful difference between your comment and mine? I can't, but I'd be very interested to hear your reasoning if you feel they are different somehow.

LLM = Large Language Model. Asking it to show its work will cause it to generate more tokens (you could think of it as the logical leaps between each step being smaller the more detail you go into)

I just paid the US State Department via check for passport renewal. Also, I expect to use my checkbook later this year to buy a car. Overall, I probably average 2 checks per year. (though I do expect that to decrease to 0 in the next 5-10 years)

One of Earth's past mass extinction events is hypothesized to have maybe been caused by a gamma ray burst[0]. I don't know enough to speculate whether it could have been caused by one of these jets instead.

[0] https://en.wikipedia.org/wiki/Late_Ordovician_mass_extinctio...

Well it seems like we've detected 3 other jets pointed straight at us as the black hole devoured a star so it seems like a very common event on a geological time scale. The fact that we managed to evolve and are still here is a decent sign.

>The team says the black hole's jet may be pointing directly toward Earth, making the signal appear brighter than if the jet were pointing in any other direction. The effect is "Doppler boosting" and is similar to the amped-up sound of a passing siren.

>AT 2022cmc is the fourth Doppler-boosted TDE ever detected and the first such event that has been observed since 2011.

Also, I believe that as the universe's volume expands, the probability and intensity of being in the direct path of any particular such jet goes down. Then again, the frequency of these events may (or may not) be increasing at a rate that more than counteracts that. (I'm just speculating here, I'm not a cosmologist!)

I can't seem to find a single site (which would presumably mean a consistent methodology) that lists recent data for average|median use per day for both of them, but I've seen numbers saying people spend 33-50 min per day on Facebook, vs 90+ min per day on TikTok.*

I suspect the curated short video format is what really holds people's attention in a way that is different from "traditional social media".

*disclaimer: as reported by different sites, with different methodologies, and different reporting dates.

Link for those who haven't read it yet:

https://scp-wiki.wikidot.com/scp-3008

The experiment you describe reminds me of a short story "What's expected of us" by my favorite author Ted Chiang. [0]

I won't spoil it too much because it's a fun short read, but it revolves around these little toys called "Predictors" which have a light that turns on exactly 1 second BEFORE you push the button. (no matter what you do to try to game it)

[0] https://www.nature.com/articles/436150a

There are costs to onboarding a new employee (not just in money, but also in time and attention from more-tenured teammates, finding good "starter" tickets for them to work on, longer standups, interviewing their replacement after they leave, etc)

For pretty much any team, there's some ramp-up time where a new employee won't be contributing much and may be taking more time/attention from your experienced employees and manager. (explaining how the system works, code reviews, helping debug why something's not working, etc etc) There's some break-even tenure below which it's a net loss for a team to have hired someone. It'll be different for each team and also different for each new-hire. Maybe it's only a couple days, or maybe it's 6+ months. Either way, managers should at least be aware of this when making hiring decisions. (exception: unless they're at a cushy enough gig where there's not much/any pressure to deliver)

There are people on HN working for the big tech companies who are being granted that $15k worth of stock every month. ($180k/year) At the more senior levels and/or with some stock appreciation, it's not unheard-of for >50% of compensation to be in the form of RSUs.

Obviously they should sell-on-vest and diversify into index funds, but a lot of them choose to buy and hold some individual stocks that they're bullish on as well.

Probably because the pilots were going to be flying an IFR approach. (flying on just instruments through low visibility near the ground, a worst-case scenario for interference being able to cause a fatal crash)

Prepaid phones all-too-often can't be used for SMS/phone authentication. Banks in particular seem to dislike them.

(when it doesn't work, you'll usually get an error message about the number not being supported or words to that effect)

Yeah, "Schemaless" is a total misnomer. You either have "schema-on-write" or "schema-on-read".

If they want to actually start and collect a few paychecks, then they'll need to pass a background check from HireRight or whoever. This'll probably include transcript checks, verifying past employment, etc.

Much easier to just use a real person's identity.

Those full bars of coverage you had at altitude are probably going to disappear the moment you lose line-of-sight.

That said, aircraft have ELTs which are supposed to trigger on impact and can be manually activated as well, so an InReach probably wouldn't make a life-or-death difference very often.

Slightly apples-to-oranges comparison on the satellite count.

Unlike Globalstar, the Iridium network uses polar orbits which means they're covering a larger area and thus need more satellites to achieve comparable service in the limited latitudes that Globalstar covers.

(that being said, if you're going to be in mountainous terrain, higher latitudes, far from land, etc then Iridium will have better coverage)

Being able to have a 2-way conversation with search & rescue to explain the situation and get updates is the biggest one.

As for other things worth paying for: being able to text your family back home and let them know you're running a day late but everything's ok so not to report you as missing goes a long way to helping everyone involved have some peace of mind.

Coordinating pickups and getting weather updates are also big value-adds.

Perhaps Amazon is FINALLY starting to do something about their rampant book piracy problem?

[0]https://mobile.twitter.com/martinkl/status/15511491594026680...

[1]https://nypost.com/2022/07/31/pirated-books-thrive-on-amazon...

Yeah, it totally blew my mind when I first heard these performance numbers as well. RE "nearly 100 million requests per second": as of Prime Day 2022, it has actually handled >100 Mreq/sec!

>Over the course of Prime Day, these sources made trillions of calls to the DynamoDB API. DynamoDB maintained high availability while delivering single-digit millisecond responses and peaking at 105.2 million requests per second. [0]

[0] https://aws.amazon.com/blogs/aws/amazon-prime-day-2022-aws-f...

No, I'll apparently have to bump it up my to-watch list!

That's exactly what I did. I had my checkbook and was there to put in a deposit when I found out. Here's an abridged version of how the conversation went:

  > Me: I don't want that and won't be paying for it.
  > Them: Well we do it for all new cars and you can't have us not do it.
  > Me: Well then you can either eat the cost of this unnecessary "service" or sell me the car for $700 under MSRP to make up for it. 
  > Them: Sorry, we can't do that.

I walked out and they haven't called me back to say they were bluffing. Meanwhile, I'm working with a different dealership now.

Hey @TreyGuy (or others from Markup.org)

How do you account for dealers which sell for $0 above MSRP but then include a non-negotiable $700 "paint protection" service?

As a crowdsourced site, you'll get a ton of reports that a particular dealership has no markup and few if any reports from people who have actually been far enough into the purchase process to discover these hidden markups.