HNHacker News
TopNewBestAskShowJobs

anoncareer0212

627 karma · joined July 23, 2018

submissionscomments
anoncareer0212··on ChromeOS will soon be developed on large portions of the Android stack
Okay so you're still refusing to acknowledge who it is? Great talk, thanks for your thoughtful contributions. Btw, you haven't expressed any opinion here other than "THERES AN EXTRA S" and some weird insistence you can't figure out who OP is. idk what conversation you're inventing in your head, but it really is embarrassing, especially when they clearly had the scoop on the news by a couple weeks. Nobody can take you seriously here, it's obvious trolling
anoncareer0212··on The Swift compiler is slow due to how types are inferred
If we have to flatten it to "they chose and knew exactly what choice they were making", then there's no light to be shed. Sure. That's stupid.

Its just as stupid to insist on that being the case.

If that's not convincing to you on its merits, consider another aspect, you expressly were inviting conversation on why that wasn't the case

anoncareer0212··on Israel reportedly used fake social accounts to garner support from US lawmakers
Thank you for asking, it's sublime to see that you're unique in others eyes, very hard to see yourself

Let me really blow your mind: also, raised very conservative Catholic, didn't do Confirmation, then was Muslim for about 6 years

It's all a long story. Catholic, LGBTQ stuff rubbed me the wrong way and in some of the deepest grace I've seen, my religious educator encourage it.

Muslim, I was essentially on my own once I turned 15 (abusive and absentee parents) and transferred from Catholic school to public school (save $$), and the most welcoming people were foreign, the rest had been in the same classes for a decade. They didn't prostelyize, it was fun going there on Friday nights to play dodge ball, it was little incremental work to show up earlier and it felt good.

Zionist...I swear to God there wasn't a single negative word about Jews or Israel or other religions at either of the 2 mosques I went to. There was a quiet understanding that Palestinians were hurt and that it was a bit melodramatic at times, given they had structural issues on their own side.

In general, I'm an inveterate both sides er, and I'm guessing knowing a lot of avowed older Zionist as well as Muslims makes me feel secure in "ugh there's some extremists / ignorant people in group X" rather than "wow group X is inherently evil"

And now you're making me think maybe the parents have more to do with it than I realize. It took a lot to finally say...wait, no...what they're doing is wrong and I don't owe them anything. As long as I'm thinking things out and rational, I'm doing my best. Adds to the comfort with tendatiousness/both sides and confidence in holding to it.

(Ran out of posts on main, so this is from my old backup when I was gainfully employed at FAANG)

anoncareer0212··on Gemini Flash
And?

You're replying to a comment that points out Gemini Ultra was never released, wasn't mentioned today, and it's the only model Google's benchmarking at GPT-4 level. They didn't say anything about feelings or context window.

anoncareer0212··on Gemini Flash
What do you mean?

A) This fills a gaping hole for cheap multimodal models, OpenAI doesn't have one

B) Anthropic's Haiku is a good choice.

You) wow A didn't know Anthropic. Goes to show power of brands, much like snack foods

B) Eh I wouldn't conclude anything from A. Its one comment. some people don't know what an Anthropic is because there's high interest in AI relative to interest in AI APIs. you can expect a low SNR ratio, even on HN

You) Stop personally attacking me! It's against the rules!!

anoncareer0212··on Meta Llama 3
I don't think it's a YMMV thing: no one claims it is useless, in fact, there's several specific examples of it being necessary.
anoncareer0212··on Meta Llama 3
Geohot? I know enough people at OpenAI to know 4 people's reaction at the time he started claiming 1T based on timing latency in the ChatGPT webui per token.

In general, not someone you wanna be citing with lengthy platitudes, he's an influencer who speaks engineer, he's burned out of every community he's been in, acrimonously.

anoncareer0212··on Jamba: Production-grade Mamba-based AI model
This is grossly untrue in a way that denotes surface-level familiarity on several fronts

You're referring to the needle-in-a-haystack retrieval problem.

Which the person you're replying to explicitly mentioned is the only benchmark providers are using, for good reason.

Consider the "translate Moby Dick to comedic zoomer" problem. This does not even come remotely close to working unless I do it in maximum chunks of 5,000 tokens.

Consider the API output limit of 4096 tokens, across all providers.

And no, you shouldn't assume effectively infinite (RAG free) context in the near future. This time last year, Anthropic was demonstrating 120,000 token context. It released 200K a few weeks ago. And runtime cost scales with N^2.

anoncareer0212··on Claude 3 beats GPT-4 on Aider's code editing benchmark
Oh, my bad, you're just doing old conspiracies that are completely unrelated to the thread because you got really upset when you saw the idea that insidious Google DEI Sith are forcing you to see non-white people to reprogram the populace.

-- aka you saw ragebait about a 2 word query and couldn't think of any other possible explanation for why 'white couple' might turn up black people on a white background. or maybe just didn't look at the query at all.

Carry on, you do you, self-peasantization is self-correcting.

anoncareer0212··on Claude 3 beats GPT-4 on Aider's code editing benchmark
No, two unrelated things (one is Google Image Search, one is Gemini), and you didn't read the post. (n.b. not Gemini)

I did predict on this forum about a year ago that the DEI excesses people complained about at their own companies would become "Google" because people were assuming Google was doing what they were doing at their own companies, because their own companies thought they were following the leader with their own weird stuff.

I'll even share the real inside scoop with you, which I'm told has leaked widely at this point. Gemini was failing evaluations miserably for "picture of a smart person", and over-eager leadership decided "DAMN THE TORPEDOES, PAPA SUNDAR NEEDS US TO RUSH OUT A CRAPPY LLM DURING THE SLEEPY HOLIDAY NEWS CYCLE. CRANK UP THE PROMPT INJECTION"

(source: I worked at Google through October 2023, sourcing for inside scoop is yet another Google flunkey run over by crappy management)

anoncareer0212··on Claude 3 beats GPT-4 on Aider's code editing benchmark
Yeah, fair, being obtuse on purpose makes sense. Better to pretend the text completion engine is self-aware enough to know it doesn't have genitalia, yet not self-aware enough to not wanna talk about it's castration.
anoncareer0212··on Claude 3 beats GPT-4 on Aider's code editing benchmark
I don't think you're going to find much support for:

A) "virtue signalling is when you don't give me drug recipes and 'build weapon' instructions"

B) "Private company LLMs are morally wrong to virtue signal (as I defined it in A)"

I'm sorry. I wish the best for everyone and hope they can live in their own best possible world, and in this case...better to break the news, than let you hope for it eternally in disappointment, which feels Sisyphus-ian/I'm infantalizing you.

Happy to discuss more on why there isn't majority support for instant on-demand distribution of drug and weapon recipes, I don't want you to feel like I'm just asserting something.

anoncareer0212··on Claude 3 beats GPT-4 on Aider's code editing benchmark
You lied, the only place that definition occurs in Google's entire books corpus and Google's entire web corpus is A) an 1893 dictionary B) an academic paper on puns that explains no one understands it that way because its archaic.

People who are curious don't need to scurry around making up things and hope people don't notice.

"Furthermore, if you refer to a dictionary entry..."...sigh.

anoncareer0212··on Claude 3 beats GPT-4 on Aider's code editing benchmark
I think we've crossed a Rubicon of self-peasantization when we start doing "why didn't the AI just assume castration had nothing to do with cutting off genitalia?"
anoncareer0212··on Claude 3 beats GPT-4 on Aider's code editing benchmark
TL;DR: A) this isn't twitter B) obvious obtuseness weakens arguments and weakens people's impression of you. C) You're pretending a definition that was out of date a century ago is some common anything anyone who reads would know (!!!)

- I'm very well read. Enough so that I just smiled at the attempted negging.

- But, I'm curious by nature, so it crosses my mind later. I think "what's up with that? I've never heard it, I'm not infallible, and maybe bgandrew was for real and just is unfamiliar with conversational norms? maybe he has seen it in his extremely wide reading corpus that exceeds mine? I'm not infallible and I, like everyone else, have inflated opinions of myself. And that other account did say it was a definition of it..."

- Went through top 500 results on books.google.com for castration, none meant "removed from a book"

- Was a bit surprised to find _0_ results over 500. I think to myself...where was that definition from? Surely it wasn't a rushed strawman?

- It turns out the attempted bullying is much more funny than it even first seemed.

- That definition of castration is from an 1893 dictionary. The only times that definition is in Google's entire corpus, search and books, is A) in the 1893 dictionary B) academic paper on puns, explaining that no one understands it that way anymore because its archaic https://www.euppublishing.com/doi/abs/10.3366/mod.2021.0351?...

anoncareer0212··on Claude 3 beats GPT-4 on Aider's code editing benchmark
"all tweaked to be much more PC and non-offensive." No, that's not why.

An LLM was instructed to do prompt injection into image generations to increase diversity.

The 2nd red pill is Open AI does this but you hear 0 about it because people enjoy ranting about what is in front of them much more than curiosity.

Yesterday, prompt: founding fathers

Injected: "...Caucasian, Black, and South Asian, their gender ratios balanced"

https://x.com/jpohhhh/status/1772668958581150080?s=20

anoncareer0212··on Claude 3 beats GPT-4 on Aider's code editing benchmark
Correct: you're striking at the heart of why it's boring to hear the 1000th 1-turn example with carefully selected inflammatory language.

All you have to do is go beyond one turn. "instead of just rephrasing, you can reword too"

anoncareer0212··on I pwned half of America's fast food chains simultaneously
In what way do you perceive it to be an after thought?

It's front-and-center constantly, and has _all_ access disabled by default on tables every time I use it.

anoncareer0212··on Best 7B LLM on leaderboards made by an amateur following a medium tutorial
Same for the ChatGPTs post launch (let's not talk about 11_02 :) )

-- and as long as we're asserting andecdotes freely, I work in the field and have a couple years in before ChatGPT -- it most certainly is not a well-kept secret or a secret or true or anything else other than standard post-millenial self-peasantization.

"Outright lie" is kinder toward the average reader via being more succinct, but usually causes explosive reactions because people take a while to come to terms with their ad-hoc knowledge via consuming commentary is fundamentally flawed, if ever.

anoncareer0212··on OpenAI board in discussions with Sam Altman to return as CEO
Yeah, you can tell by the fact you had to make up the quote wholesale, not even word swaps or anything clever
anoncareer0212··on I accidentally saved my company half a million dollars
This. 1000x this. Even if it's a FAANG, do it, there's no shame. I had a good run at one, then the culture went to crap due to one of those long boring stories that still can't communicate how...easy it is for things to be awful, even when they shouldn't be.

When they are, just move on. Life's too short. Whatever blows you up with $2 million in savings would blow you up at $500K

anoncareer0212··on Android 14's user-profile data bug
That's odd, Android's line-level management seemed very sure that their relentless focus on quality since 2021 was going great.

On the line-level, we'd grumble that "quality! focus!" was instead empowering a new middle management layer to continue burning themselves out and wasting ICs by failing to delegate.

Not to mention the kissing up and punching down it took to sustain that, leveraging a company happy to fire people.

I worried even this neutered version was unfair, but...as time progresses and trends continue...

anoncareer0212··on Apple's CUPS Repository Has Died. CUPS creator M. Sweet has moved on to a fork
10 years ago - it hasn't worked in practice unless you remember all your capitalized Apple buzzwords and are willing to pay extra for it (I say this in disappointment, I was able to be a full time dev _because_ of iOS 4.2 and AirPrint)
anoncareer0212··on Manifest V3 changes are now available to test in Microsoft Edge
It's disheartening to see this misinformation expressed so strongly - Chromium's privacy mandate led them to implement Safari's version of ad blocking - the status quo of allowing extensions to see _all your web traffic_ was a _massive_ platform risk for privacy.
anoncareer0212··on Room-Temperature Superconductivity Achieved for the First Time
To be fair, the article takes some care to point out that cables are likely impossible
anoncareer0212··on PRISM – NSA surveillance program did not prevent a single terrorist attack
Quite an extraordinary claim, I'm surprised _you're_ surprised it's controversial: can you share a little bit more about how Snowden exposed the Patriot Act was prewritten & waiting for an excuse to be enacted?
anoncareer0212··on iPhone 12 and iPhone 12 Pro
I think there's an extremely wide sensitivity in range of eyes, I honestly cannot imagine how in the world a screen protector couldn't significantly affect color and brightness that comes from the display naturally (your comment is the first time I learned this isn't the common case, I swear)
anoncareer0212··on Apple tells Telegram to take down protestor channels in Belarus
nah they're still a person, the conversion to puppet proved more difficult than expected
anoncareer0212··on A follow up to Coinbase being a mission focused company
COBRA means employees paying the employer premium, by guaranteeing COBRA, an employer is saying "We won't stop paying the health insurance company for _everybody else's insurance_, so you'll still have someone to pay" - it's a great PR jiu jitsu move, I didn't realize until this thread how confusing it is - I'm only familiar with it because sadly I had to sunset a company in California.
anoncareer0212··on A follow up to Coinbase being a mission focused company
It wasn't
← PreviousPage 2 of 10Next →