HNHacker News
TopNewBestAskShowJobs

jeffparsons

2,340 karma · joined January 3, 2019

submissionscomments
jeffparsons··on Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario
It depends on there being _some chance_ that it can be achieved.

The benchmark strategy we are comparing it against is tantamount to licking a wall repeatedly and expecting that to achieve something. So even a strategy that may well fail but has _some_ chance of success is a lot better by default.

jeffparsons··on Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario
Sure, I think that kind of "coordination" is a little more credible than what I thought Eliezer was proposing. Maybe I just didn't read carefully enough.

One big difference here, however, is that the barrier to entry is lower. You can start doing meaningful AI R&D on existing hardware with _one_ clever person. The same was never true for nuclear weapons, so the pool of even potential players was pretty shallow for a long time, and therefore easier to control.

jeffparsons··on Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario
It can't be the only card left to play. There are so many others. They may be weak, but they still have to be better than one that is guaranteed to be absolutely useless.

Example of a weak, but at least _better_ option: convince _one_ government (e.g. USA) that it is in their interest to massively fund an effort to (1) develop an AI that is compliant to our wishes and can dominate all other AIs, and (2) actively sabotage competing efforts.

Governments have done much the same in the past with, e.g. conventional weapons, nuclear weapons, and cryptography, with varying levels of success.

If we're all dead anyway otherwise, then I don't see how that can possibly be a worse card.

Edit: or even try to convince _all_ governments to do this so that they come out on top. At least then there's a greater chance that the bleeding edge of this tech will be under the stewardship of a deliberate attempt for a country to dominate the world, rather than some bored kid who happens to stumble upon the recipe for global paperclips.

jeffparsons··on Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario
My estimation of Eliezer Yudkowsky just dropped a notch — not because I don't take the existential threat posed by AI seriously — but because he seems to think that global coordination and control is an even remotely possible strategy.
jeffparsons··on Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario
Which would still mean that those who can afford more/better hardware can therefore afford more/better AI. I don't think a solution lies down this path...
jeffparsons··on Help Wanted at LWN
IMO good tech writers are way undervalued in general, so I'm really happy to see that this position appears to be well-compensated. If I were an LWN subscriber, I'd also be happy to know that my subscription is contributing to that person's salary.

I still remember when Steve Klabnik moved on from Mozilla he mentioned in a blog post [1] that he'd recently discovered he was the lowest-paid member of his team. I was simultaneously flabbergasted, because I felt that Steve's contribution to the Rust project was so huge that it was hard to imagine what the Rust experience would have been without him — and entirely unsurprised, because it's so typical of organisations to undervalue writing work.

Writing is communication. In my experience, in any endeavour involving more than one person, communication is the single biggest factor in success or failure. So... you know what I'm getting at: pay your communication specialists well.

[1] https://steveklabnik.com/writing/thank-u-next

jeffparsons··on King Charles III signs off on Online Safety Act with unenforceable spying clause
The monarchy might not last very long if it were to actually exercise much of its theoretical power. UK folks love their royalty these days as a kind of mascot, but I doubt they'd tolerate an interventionist king.
jeffparsons··on Windows 10 warns me to use a "Microsoft-verified" app
Just look at Uber, for example. The opportunity cost of obeying the law in the various markets they entered would have far exceeded the cost of penalties they received in practice. So what did they do? They just flat out ignored the law globally, and it seems to have worked out pretty well for them.

Unless part of the calculus involves executives behind bars, I don't know what else we should expect from corporate behaviour; they're just following their incentives.

jeffparsons··on Scrollbars are becoming a problem
I might be getting old, too. (Born 1987.) I have also come to regard Windows XP as the pinnacle of personal computing UI. Almost everything that followed, regardless of vendor, feels like a step in the wrong direction.
jeffparsons··on Desmos 3D graphing calculator
If you appreciate quaternions for what they can do for you, but also find them frustratingly unintuitive/ugly, then you may be interested in bivectors (https://bivector.net/) and geometric algebra more broadly. They're equivalent in power, but suddenly everything actually makes sense!
jeffparsons··on Divan: Fast and Simple Benchmarking for Rust
The recently-formed testing subteam is looking to solve your exact complaint: https://github.com/rust-lang/rfcs/pull/3455
jeffparsons··on Goodbye integers, hello UUIDv7
Yeah, you can require them to generate a UUID if it helps for your app (e.g. bulk create objects with relationships between them) but then on the server you can generate new UUIDs and return a mapping from old->new to the client.
jeffparsons··on The State of Async Rust: Runtimes
> Wasm has been really great at shipping the MVP, but they are pretty slow about shipping the many features that build on it. In general, this makes sense as the system can't be changed once its stable. But it also means that a lot of things are still in limbo and will probably be for the forseeable future.

I concede that development of post-MVP development has been slow, but I am also optimistic for the near future based on recent progress. Are there particular things in limbo that you're particularly interested in or concerned about?

jeffparsons··on Collective letter to unity from game development companies
I'd bet that most workers would benefit from more unionisation (including those who think they're above the need for that sort of thing — think software engineers), and most small businesses would benefit from more cooperation to protect them collectively from the giants.

But organising people is _hard_ and so we tend to give up a lot of control for a little convenience. Unity can pull the rug at any time, and there's not a damn thing you can do about it. However, if 1000s of independent game developers had a share in a similar company whose constitution dictated that it was for the benefit of those independent game developers, and they could vote out directors that allowed it to deviate from that purpose, then this scenario we're seeing today would be a lot less likely.

jeffparsons··on The State of Async Rust: Runtimes
From my occasional skimming of WebAssembly meeting minutes, I'd say that Wasm will likely grow the features required for Go to perform well. There's plenty of interest in stack switching, coroutines, etc.

I work a lot more in Rust than I do in go, but I think each language made the trade-off that made most sense for that language.

jeffparsons··on Programming Language Inventor or Serial Killer? (2003)
I think it's because it looks like an _ancient_ security camera feed, which reminds me of watching security footage on television news as a child. Most modern systems I see don't have that kind of noisy distorted "bad analogue" look so I think they just don't trigger the same reaction in me.
jeffparsons··on Programming Language Inventor or Serial Killer? (2003)
There's a security camera monitor displayed prominently in one store near where I live that makes everyone look dodgy — something about the frame rate and fuzzy image just looks like shoplifting footage you'd see on the news.

I see myself on the screen as I enter and think "yeah, that guy's going to hold up the shop or something".

jeffparsons··on Reaching the Unix philosophy's logical extreme with WebAssembly
To play devil's advocate: targeting WASIX still gives you the benefit of compiling existing code once and running on any OS-architecture pair that Wasmer supports. I long for this sort of thing when building software that needs to run on Linux on multiple architectures inside AWS and also on developers' MacBooks. (Ever tried cross-compiling to macOS? Yeah, I prefer pulling teeth, too.)

Caveat: you'd need to use Wasmer forks of, e.g., rustc and Tokio, but I imagine there are at least some people who would be okay with that.

jeffparsons··on Bigger, better, faster, more
I'm a very happy NFS customer. I give them money, and they host stuff for me. If I run out of money in my NFS account, they stop hosting stuff for me. Simple, fair, low-risk. (There are alarms, "suspended mode", etc., so they don't just pull the plug, either.)

My favorite feature of this way of doing things is that I never need to wonder how much they're going to bill me, because I've already pre-paid the maximum I'm willing to lose in one go should I fat-finger something. Compare this to the unlimited liability involved in using AWS et al., which should be a terrifying prospect for most individuals.

jeffparsons··on Apple clarifies why it abandoned plan to detect CSAM in iCloud photos
I feel this neatly captures the overarching corporate philosophies and attitudes of Apple and Google in a single example.

Pick your favourite other example of when Apple and Google have faced roughly the same problem as each other, and hold up their respective solutions next to those in the example of CSAM scanning above. I bet they'll look similar.

jeffparsons··on Downloading a video should be “fair use” as recording a song from the radio
I'd love to believe this, but said regulators don't seem particularly energetic these days.
jeffparsons··on PackagingCon – A conference only for software package management
Definitely. Much of this can be mitigated by using a shared target directory, which is at least somewhat supported in Cargo. I should probably start doing that and promote it at work if I don't run into any problems.

Some clever garbage collection would help, too, but I imagine different people would have different and very strong opinions about how that should work.

The fact that two projects that use the same library may enable different features also complicates things. Again, a of this can be mitigated.

I'm looking for a step change for application development, though — i.e. not 5 minute build reduced to 2 minutes, but 5 minutes reduced to 5 seconds (and a correspondingly tiny target directory). That's what excites me about WASI in this context at least.

Leaving WASI aside for a minute, I do wonder how much more could be saved in local disk space and compilation time across projects (and hosts, a la sccache) if this was a high priority goal for the Rust project. E.g. even if the MIR for a crate with two different sets of feature flags enabled ends up substantially different, would they still compress well against each other if a lot remained common?

jeffparsons··on PackagingCon – A conference only for software package management
I'm sure we can all agree that Cargo is the best thing since sliced arrays, but I've finally reached the point where we have enough different Rust projects at work that rebuilding the world from scratch all the time and having ~40 GiB `target/` directories scattered around the place is getting old.

I'm not meaning to criticize Rust or Cargo here; they're both doing their jobs just fine. But I do find myself craving a different compromise for much of what I/we use Rust for today. And I'm really hoping that Wasm Components (and WASI) will be that different compromise — e.g.:

- Don't rebuild an entire HTTP stack from scratch for every tool that happens to use HTTP. - Therefore most projects won't have enormous `target/` dirs. - Reuse components built in CI for Linux directly on a Mac dev machine (cross platform). - Mix and match new processor architectures available in AWS without building two versions of everything. - Also reuse some components in the web browser (yes, we actually have real, boring use cases for this). - Wait, I'll be able to do all of this and still be writing Rust? Shut up and take my money.

jeffparsons··on WASI: WebAssembly System Interface
I imagine Wasmer must be in a bit of a difficult place right now. My understanding is that they are VC-backed (YC?), which means they're essentially trying to figure out how to make a huge amount of money off an open standard or something adjacent to it. But then, they would have known this from the very start, so I'm not sure what they expected, or why a VC firm would even look at them as a promising prospect.

Maybe they hoped to be like Docker, where the value-add was cohesive higher-level tooling on top of all the lower-level bits. If that's the case, then WASI would be an existential threat in that it (and the tooling people are building around it) commoditizes what would otherwise be Wasmer's special sauce. I'm speculating wildly, of course, but it would at least be consistent with Wasmer's/Sirius's apparent vendetta against WASI and the people working on it.

I hope Wasmer eventually finds a way to add value that doesn't rely on constant gaslighting and white-anting the standards process.

jeffparsons··on WASI: WebAssembly System Interface
Genuinely curious: what are you working on that makes POSIX compatibility important, but also warrants using Wasm instead of just building native executables?

I ask because I'm personally much more excited about the future of WASI, because it brings so much to the table that it feels like it's actually worth the effort of building something new — I see it as a game changer. I.e. what you appear to view as an unnecessary diversion, I see as the whole point. So I'd love to understand more about your perspective and use case.

jeffparsons··on WASI: WebAssembly System Interface
> version 1

You mean preview 1? It was never meant to be a finished spec that would remain supported long-term; it was a preview for early adopters and people working in the space to experiment with.

> allowed to continue for political reasons

Uh... what?

jeffparsons··on Bun v0.7.0
I always saw it as a xiaolongbao specifically. Maybe that's my personal preference biasing me, though. (And now I'm hungry.)
jeffparsons··on My broken ThinkPad plays music upon booting
Where I am, two beeps means concession fare, so the driver/whoever can notice and ask to verify your concession entitlement. Might that have been it?
jeffparsons··on PostgreSQL: No More Vacuum, No More Bloat
I think I basically agree with you. And this example is pretty benign — I'm not actually meaning to criticize anyone here.

However I will not that the author in question refers to himself as "Rick Branson", and the article title is "10 Things I Hate About PostgreSQL". So I think it's just the person who made the link who is being a bit cheeky.

My comment was going off on a wild tangent. :)

jeffparsons··on PostgreSQL: No More Vacuum, No More Bloat
True words can still be clickbait.

In fact, I'd argue that many of the most effective ways to mislead people involve sticking rigidly to literal truth, because it makes them so much harder to counter. When there's no literal untruth to correct, it's natural to end up implying bad faith _without having any definitive proof_, and that is mighty unstable ground from which to argue.

← PreviousPage 5 of 12Next →