- > Probably not. All side effects need to go through the js side. So you can alway see where http calls are made
That can be circumnavigated by bundling the conversations into one POST to an API endpoint, along with a few hundred calls to several dummy endpoints to muddy the waters. Bonus points if you can make it look like an normal-passing update script.
It'll still show up in the end, but at this point your main goal is to delay the discovery as much as you can.
- > > For a perfectly spherical person
> The transformation of humans to spherical shape is gluttony not evolution. They should be forced to ride a bike until their silhouette returns to that of a human.
...GP's comment is a play on the "spherical cow" physics joke, and how models *will* have some unrealistic assumptions baked in, just so that the maths is easier to crunch through.
- Counterpoint: Such a market *technically* already existed outside of FIFA, just that it was a more underground/grey/black market.
Strictly speaking, an external market being brought into existing ticketing systems would be net-neutral, since the following pros & cons should balance each other out:
(additional visibility into ticket prices & demand (+)) + (increased assurance of "this is the one place to get a ticket" (+))
==
-( (increased competition for a ticket (-)) + (perverse incentives of platform to increase ticket prices (-)) )
But because of their reputation, the negatives are weighed more than the positives due to their existing track record.
As such, the following constructed scenario should be considered: If it was a fully automated platform external to any party that handled such ticketing systems, would such a severely negative view still hold?
- https://www.hackerneue.com/item?id=45529810
https://cpufun.substack.com/i/32474663/notable-differences
As noted by the other comments, Apple's M-series chips seem to use a 128-byte cache line. ARM doesn't mandate that their licensees must use a pre-specified cache line size: 64 bytes just happens to be the consensus-arrived standard.
- > Any reason to believe Google's unit economics on AI are any different than the other players here?
Only when it comes to their TPUs, and sometimes that one thing may just be the difference to push them over the hump.
Per-token cost-wise, TPUs (& specialized processors in general) will beat GPUs every time. The efficiency difference between the 2 types is never to be ignored, & is likely why they can shotgun it everywhere.
> And Google is an advertising company. Mostly in search, and increasingly dependent on YouTube. Everything else is a net money loser, including Waymo, Gemini etc.
1) Each venture should be treated as a (relatively) isolated vertical slice
2) 9 out of 10 times, a venture just doesn't break even. That's just the nature of the business.
- Mandatory xkcd link: https://xkcd.com/297/
Related performance-oriented discussion: https://www.hackerneue.com/item?id=40296932
Lisp will always be the 'what if' watering hole in the compsci space.
- > Tracking and restricting the freedom of criminals is critical to a healthy society.
Under such an excuse, any type of extreme measure is similarly allowed.
In fact, why not go full Suicide Squad and strap bomb/shock collars to every person on Earth?
- This is just a benchmark of how much of a sycophant an LLM is. Anything that scores > 50% on this test should be punted into the bin.
- 7000000 reels ÷ 50 years ÷ 365.25 (days/year) = ~383.3 reels / day
Each reel being a minute long would equal to 6 hours & 24 minutes of scrolling a day.
It would be close to the most depressing world record to ever exist.
- I will always push Dylan Beattie's talk about this exact thing!
- > Isn’t that basically what the ARC tests are?
Reductively, yes.
IMO, the ARC tests & the visual pattern IQ tests (e.g. Raven's) have little difference, especially if the Raven tests require the taker to draw out the answer.
https://en.wikipedia.org/wiki/Raven%27s_Progressive_Matrices
- > So, an ai generated psuedo-game engine with a majority of users under the age of 13? I'm sure that WILL make a lot of money. Those of us who didn't grow up playing Roblox will find this comparison impossibly stupid.
> ...with a majority of users under the age of 13? I'm sure that WILL make a lot of money. > ... will find this comparison impossibly stupid.
I'm ignoring the insinuations here for obvious reasons.
1. Roblox is the newest (note: not necessarily the best) iteration of the genre that Secondlife & (to a limited extent) modded Minecraft servers occupy: An interactive 3D platform that permits user-generated content.
2. Generative models just accelerate their development up to the brick wall of complexity much faster.
> Some what related: im still amazed that no one has made a Roblox competitor
This comment is just the HN Dropbox phenomenon, *again*, only this time from the angle that thinks it's easy to build a "pseudo game-engine" from scratch.
https://www.hackerneue.com/item?id=8863
Few competitors exist because of the moat they have built in making their platform easy to develop on, so much so that kids can use them with little issue.
> , as in, a vague social building game that tricks children into wasting money on ridiculous MTXs.
This part is entirely separate from the technical aspects of the platform. Roblox is a feces-covered silver bar, but the silver bar (their game platform) still exists.
> Maybe you are right, but I think that taking an already sorry state of affairs, and then removing the only imagination or STEM skills required by giving children access to GenAI.... is a really depressing thought.
This is a hyper-nihilistic opinion on children laid bare.
To think that the children (*with the dedication to make a game in the first place*) wouldn't try to learn about debugging the code that the models are spitting out, or that 100% of them would just stop writing their own code entirely, is a cynical viewpoint not worth any attention.
- I actually laughed at the last one.... Jesus.....
- > The question is, try to spend $1bn on stuff. Go.
Easy: The largest ship in the world by area. (Goal - either 500m x 500m, or at least 0.25km^2 with the breadth >= 300m)
The current status quo for bulk carriers are the Valemax ships (360m x 65m), with each one costing around $100 million. (actual figure wildly varies, but sticks around that number)
https://en.wikipedia.org/wiki/Valemax#Sale_of_ships
https://www.tradewindsnews.com/containerships/evergreen-adds...
(500 * 500) / (360 * 65) = 10.683760683760683
10.68 * $100 million = $1.068 billion
Even just going with 5 Valemax ships side-by-side (360m x 325m) costs half a billion.
- The section is not without its own flaws, mainly in 5(c):
> (c) Civil penalty.—Any payment card network that violates subsection (b) shall be assessed a civil penalty by the Comptroller of the Currency of not more than 10 percent of the value of the services or products described in that subsection, not to exceed $10,000 per violation.
I see 2 problems as it is currently written: (1) The penalty's too low, & (2) restricting dispense of the law to only the Comptroller renders it ineffective.
(1) is easily solvable with regards to editing the text alone: raise the limit to 50% & $100k respectively.
(2) is also solvable, by striking out "by the Comptroller of the Currency", or adding in ", or by a federal court, whichever penalty is higher, " at the end of that part.
- You're rationally thinking about this, not the impulsive state someone will be in when impulse buying (where math rarely occurs, if ever).
Visually showing alternatives to the impulse buyer adds friction between the user & the purchase, and when compared to the original seamless shopping experience, that's enough to slow down impulse purchases.
- The parent comment is an ad hominem attack, plain & simple.
Nothing has been said by the parent about the message: Instead, the messenger is marked as the target in order to take down the message via association.
"Give me the man and I will give you the case against him"
------
About the legislation itself:
https://www.congress.gov/bill/119th-congress/senate-bill/401...
It's mainly 3 semi-related parts stapled together. If the person/business/bank is complying with the law:
- 1a) Banks with > $10 billion are blocked from (i) the Fed's discount window lending program, & (ii) the Automated Clearing House Network, if they refuse to do business with them,
- 1b) Banks are required to accept deposits from them & law-complying member banks,
- 1c) The board of directors of banks/credit unions have to notify the state/fed if said access blocking happens,
- 2) (Section 5) Payment networks can't refuse service to them, &
https://www.congress.gov/bill/119th-congress/senate-bill/401...
- 3) Banks & payment networks have to give them access to financial services
The middle part (2) is what's being focused on in the article.
I'm mostly in agreement that (1) the penalty's too low, & (2) restricting dispense of the law to only the Comptroller renders it ineffective.
(1) is easily solvable with regards to editing the text alone: raise the limit to 50% & $100k respectively.
(2) is also solvable, by striking out "by the Comptroller of the Currency".
- ...It took me writing out my original question to recognize why X:=(an in-place upgrade from RAIDZ1 to RAIDZ2) is non-trivial to implement.
(Original question: Why X hasn't been implemented yet?)
https://www.raidz-calculator.com/raidz-types-reference.aspx
The general difference between the two is RAIDZ2's additional blocks from its 2nd parity function, which seems independent from the 1st.
The (ideal) set of operations that would happen would be:
1) Treat existing parity disk as blank without overwriting it
2) Write to chosen parity disks 1 & 2 from new disks (assertion: new disks >= 2)
3) Blank out old parity disk
Step (1) would be near-instant, since it involves no writes. (3) either involves removing the block entries on the old parity disk and treating it as blank (near-instant), or writing zeroes to the disk (lengthy). The bulk of the time would be spent on (2).
The process of doing this migration would require siloing off writes to another place whilst the new parity disks are being built.
Considering that this migration (A) very rarely occurs, and (B) would require a write silo in the interim, devoting time into creating this error-prone task would not be wise.
- > Why do you feel the need to be the devils advocate? What do you get from defending this crap?
To prevent this level of monopolar partisanship.
Is this action a net-negative for ISP subscribers (i.e. everyone)? Yes.
However, this doesn't give any leeway at all in ignoring the past failures of the other side when it comes to this space either. The only (legally advisable) path left is vocal advocacy for its restoration. Doesn't matter who it is, only that this goal is to be achieved.
---
Also: https://news.ycombinator.com/newsguidelines.html
> When disagreeing, please reply to the argument instead of calling names. "That is idiotic; 1 + 1 is 2, not 3" can be shortened to "1 + 1 is 2, not 3."
> Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something.
- > On the other hand, stable-coins suffer the same problems as visa. They're centralised, and subject to zealous regulations.
Not all stablecoins are the same. There exists 2 main categories: Fiat-backed (as initially described), & collateral-backed.
(There also exists hybrid versions, but they're a combination of the 2, and as such will be covered by just mashing the 2 categories together.)
Fiat-backed stablecoins (USDC, USDT) are centralized: Their connection to external cash/bonds requires them to have an accountable name to be attached to.
Collateral-backed stablecoins (GHO, DAI/USDS) don't have to be centralized. A primitive form of this is a stablecoin (S) that can take in any other token as collateral and return $X amount of S stablecoins, up to a limit of (total_token_value * collateral_limit). However, it is known that this structure is inefficient capital-wise, when compared to fiat-backed stablecoins.
- Previous discussion (520 comments): https://www.hackerneue.com/item?id=30911598
- > Capitalism, plain and simple. I don't like it either, but that's the way all fundraising has devolved down to.
No.
*Even under a socialist framework*, the tactics used for fundraising would not be alien. The main difference would be to court either (a) the public votes for funding approval, or (b) the leadership votes that hold the administration's purse strings.
Don't blame it on the system. There are only so many ways to gather resources before the different methods overlap with each other.
> In any decent world, governments would use tax money to both fund projects like Wikimedia and help get poor countries off on a self-sustaining economy, but that ship has sailed I am afraid.
Counterpoint: Why is it *my* burden to bear on building up *their* economy/business/organization? They're not as inept as implied in your saviour complex.
- The article assumes that there will be no architectural improvements / migrations in the future, & that Sparse MoE will always stay. Not a great foundation to build upon.
Personally, I'm rooting for RWKV / Mamba2 to pull through, somehow. There's been some work done to increase their reasoning depths, but transformers still beat them without much effort.
- > I needed 1 database per tenant
...Depending on how many tenants that is, Turso may fit your use case.
- > > if you just turn javascript off that gets rid of 99%
> Given how websites are built these days, if you just turn javascript off, half of them, if not more, will become unusable.
Basically any webapp with any amount of processing being done on the device becomes unusable if JS is disabled. Photopea's a good example of this.
- > In this case the woman is obviously mostly writing this because she doesn't like OpenAI specifically, but she has an argument and it is a good one.
Until the same measures are uniformly taken with the rest of the actors in the industry, this is nothing more than virtue signalling.
Denouncing X should also mean denouncing actors close to X, for a given domain space F(actor).
- > In this context, never. Especially because the parent knows you will always ask 2+2 and can just teach the child to say “four” as their first and only word. You’ll be on to them, too.
On the assumption that you'll always only ask it "what's 2+2?" Keywords being "always" & "you".
In aggregate, the set of questions will continuously expand as a non-zero percentage of people will ask new questions. The set of questions asked will continue to expand, and the LLMs will continue to be trained to fill in the last 20%.
Even under the best interpretations, this is the detractors continuously moving goalposts, because the last 20% will never be filled: New tasks will continuously be found, and critics will point to them as "oh, see, they can't do that". By the time that the LLMs can do those tasks, the goalpost will be moved to a new point and they'll continue to be hypocrites.
------
> > At what point would you be impressed by a human being if you asked it to help you with a task every 6 months from birth until it was 30 years old?
Taking GP's question seriously:
When a task consisting of more than 20 non-decomposable (atomic) sub-tasks is completed above 1 standard deviation of the human average in that given task. (much more likely)
OR
When an advancement is made in a field by that person. (statistically much rarer)
- ... It might be a pursuit worth making a small library for.
- > Yes, and it belongs to us. It's not theirs to sell to the highest bidder.
Your attention belongs to you, until you give it to someone else.
The videographer has the right to sell sponsorships on their videos in exchange the attention they've attracted. It is also their right to do so.
> Not our problem. Business needs do not excuse it. Let all those so called innovators find a way to make it without an attention economy. Let them go bankrupt if they can't.
Your logic has already been tried: It's called Netflix. And it was overtaken by YouTube.
YouTube has been the wellspring for indie videographers because they have a platform that could (a) handle the video hosting for them for free, where (b) they could post their experiments on without an upfront cost & where an audience can be found because the platform's free.
Your idea seeks upfront payment, which increases the risk cost dramatically from 0 to a fixed value. One-shot experiments with 0 funds are killed under your scheme.
To seek their bankruptcy is nothing short of a fetishistic desire for your ideals to trample on others your your own gloating. Go back to the DVD era.
https://exe.dev/docs/pricing