HNHacker News
TopNewBestAskShowJobs

jl2718

3,462 karma · joined February 17, 2018

submissionscomments
jl2718··on ChatGPT took their jobs. Now they walk dogs and fix air conditioners
How can you internet in {current year}, and seriously think we need more content writers? We’re not talking about top-tier investigative journalists blowing scandals open and speaking truth to power. We’re talking about the meaningless banter filling space between ads that turn a simple recipe search into a dystopian nightmare.

I hope and trust that they will find meaning in a career fixing and building and caring for physical things that people actually want to pay for. This is a step forward.

jl2718··on Google Photos AI still can't label gorillas after racist errors
There is a rigorous definition of bias, and that isn’t it. That is called ‘variance’. It doesn’t matter if there is structure to the variance. It doesn’t matter if the model is consistently wrong for one type of input. It is not bias unless it has been specifically and intentionally programmed to predict something other than the result from the maximum likelihood estimator given the data. A trickier type of bias is in the dataset itself, but once again, this would require a specific effort and intention to produce that bias by modifying the collection or filtering procedure. Nothing as much has been alleged, so you have provided no evidence of bias, ‘systemic’ or otherwise.

Come up with a different word. Maybe “non-uniform variance”.

jl2718··on GDP is the wrong tool for measuring what matters (2020)
Forgive me if I missed the objective quantitative measurement the article suggests to replace it.

GDP is kind of like measuring spending in your family when you should be measuring saving. It’s not meant to track personal well-being, but rather the strength of an economy, and for that, a better metric may be savings and investment rate.

Savings rate is a measurement of productivity normalized to consumption. Closely related is investment rate, which is savings times capital efficiency.

jl2718··on GDP is the wrong tool for measuring what matters (2020)
You forgot taxes. 500$ each.
jl2718··on Ask HN: My GPT project has become a local hit but I can't afford the bill
People respond well to financial transparency. If you have a donation button, show real-time stats on the amount it costs, and the amount that’s been donated.

Also because internet, expect 90% of cost to be from abuse, and likewise 90% of your dev time devoted to fighting it.

jl2718··on TV doctors say annual checkups save lives – real doctors call bullshit (2016)
What % of that cost was the zero-value-added gatekeeper?

Here is a list of prices: https://dhhr.wv.gov/bms/FEES/Documents/Clinical%20Diagnostic...

Testosterone is $46.

Toxoplasmosis is a shockingly common and easily-curable brain disease: $12.

Imagine, an entire population held hostage to this racket.

jl2718··on TV doctors say annual checkups save lives – real doctors call bullshit (2016)
There’s a website that can pass the test to become a doctor now. Why can’t I order my own diagnostics?
jl2718··on Downtown San Francisco is at a tipping-point
Such statements confuse me. Let’s go point by point on this:

1. How does one “buy” a government? With bond purchases? With campaign contributions? With voter fraud? And in the chosen case, what powers would they gain to “fix their habitat”?

2. Which valley “gazillionaire” is actively spending money in any way that doesn’t align with progressive politics and larger stronger less-accountable government? For the purposes of this question, anybody invested in Palantir is ineligible.

3. What “their way” would a strong government be “in”? What specific harm to whom are you suggesting?

4. If they have infinite power over government, why do you think they would be happy to have it “out of their way” versus active assistance?

5. If we observe that “the government” is quite powerful at the present moment, then how do you resolve the contradiction. Are they wasting their money?

jl2718··on Nvidia’s Blowout Forecast Sparks $260B AI Rally
I think there is a heavy quantitative induction mechanism in the markets, that is, a positive feedback mechanism from index funds trying to minimize divergence, or hedge funds trying to font-run those algorithms. So it doesn’t take much to set off irrationality, and not much exists to pull it back, because income investing basically doesn’t exist, and therefore value investing has little support.

Partial side note: obviously the FOMC has a lot to do with investment flows, but I’ve found that the tax system is just as relevant or more so in market distortions, but it is poorly understood, rarely observed, and never considered for alternative possibility, like the water that our economy swims in. For instance, the combination of 401k and double taxation killed income investing. Also, capital gains tax and long-term preference deters sale in a pump, and actually works the other way to prefer short-term selling of losses, so the net effect is to add to the positive feedback. Every tax situation is a comparison of speculative gain and loss on an efficient market hypothesis prior, versus a guaranteed tax loss.

jl2718··on Ethereum Ceremony – Add your randomness to prepare the way for EIP-4844
“Danksharding” is named for a person, “Dankrad”, who has been their expert and champion of elliptic pairing-based cryptographic mechanisms.
jl2718··on AI ‘Cheating’ Is More Bewildering Than Professors Imagined
Right. My high school rhetoric teacher used to label essays as “garbage” for not following structure or convention, often even for straying from expected narrative, while I was only interested in forming a unique or strong perspective. We may have both been wrong. I think the future of “essay” writing will be sort of like patents in that they will have to first identify a problem with a well-established viewpoint, then propose a solution, and fully describe its novelty by citing the prior works that form the boundary of the idea.
jl2718··on Students give Gilbert Strang a standing ovation after his last lecture
He taught this class long ago at Lincoln Laboratory, tests, problem sets and all, to a bunch of adult learners, often science and engineering PhDs, that realized the importance of the subject only after leaving school. It transformed the culture of the place because everybody was speaking the same language across many disciplines.
jl2718··on AI ‘Cheating’ Is More Bewildering Than Professors Imagined
The AI-generated essay will be the new baseline. If you can’t argue beyond what an AI can or will generate, you haven’t broken through the NPC threshold.
jl2718··on The promise of crypto has not lived up to its initial excitement
> crypto

Please stop using this prefix by itself.

jl2718··on Ask HN: Is YC becoming less of a home for hackers and more for MBA-types?
$500k is not just enough to hire builders, it’s way too much for a builder. They’d be forced to do everything but build. In the beginning, it was just enough for a team to survive the summer, and that seemed generous.

The tech startup formula has merged with the startup hedge fund or VC fund formula. It has mostly to do with connections to capital. I’d also suggest that most ‘tech’ startups have nothing to do with technology. The biggest have been weird schemes to give something away for free or run unlicensed marketplaces.

At some point we have to admit that technology skills are a tiny part of the formula, but still worshipped because it’s the only part of it that is accessible to all, sort of like how we respect athletes more than sports broadcasters. MMA had a parallel rise, so it’s kind of like YC started out developing fighters and now they start dojos.

Can a fighter start a dojo? Yes; that is the natural order of things, and many do, but many more dojos are started by people with more money than skill.

Startups were not new in 2005; the fervor was just as strong as today, but the role of money was a lot less. The standard advice was ‘going Bedouin’, which worked because the cost of living was a lot lower, and it was much faster and easier to get a job if you needed one. Most people funded from job savings, part-time work, or living with parents[1]. The Facebook deal was unprecedented and weird. Even he took some very strange deals to fund his ideas before that came along. I know because I had a ‘job interview’ with a guy that MZ had some weird deal with to co-fund their projects, of which his was a parking app, and he was mad at ‘the Harvard kid’ for taking $6000 and not doing the work, which he wanted me to finish. I was working on a ride sharing app, so he offered me the same deal. He went to jail soon after, and then sued for 25% of Facebook when he got out. There’s also the infamous deal with ConnectU. All I’m trying to say is that everybody was hustling in weird ways.

Now it seems that the money is freely available if you are in the flow of it, and no amount of sheer brilliance can compete with that if you’re not. I think that may change in the near future, but I’ve been saying that for over a decade.

Also, if I may editorialize even further, the standard advice has always been inconsistent or inapplicable broadly. I liked the YC idea, but didn’t even apply because I was literally taking his advice not to focus on things like that, but obviously club membership has had benefits. When you succeed doing it your way, hacker, hustler, MBA, YC, vagabond, or whatever, you can tell everybody that’s the only way to do it.

[1] http://www.paulgraham.com/startupfunding.html

jl2718··on Widely used chemical strongly linked to Parkinson’s disease
Smells like fodder for a lawsuit that benefits nobody but lawyers and bureaucrats. Frankly, there’s nobody in the system who has any incentive to defend the taxpayers against these things. Imagine what it would be like if they could sue any organization for any disease prevalence that popped hot on a t-test for any period of time. And they had access to everybody’s medical records to do that. And the organization doesn’t even have to pay. It’s a nonprofit, and the people that have to pay are the original donors, again. But these weren’t willing donors; they were forced to ‘donate’ by people with guns. Oh, and they were also forced to pay for the study that is now forcing them to pay again for the result and of course more studies.

Torts against the government are a funny thing. It requires anti-government attitude to suggest that the government is incompetent, harmful, or criminal, but the solution is always more programs, more taxes, more laws, more bureaucracy. Oh and the ‘victims’? They will sign on expecting a huge payout, claim all kinds of problems, and end up with $0 and a weekly appointment with a VA occupational therapist. If they happen to have a problem that can actually be solved, too bad, no funding.

jl2718··on Amnesty International criticised for using AI-generated images
A good rule of thumb is that people or institutions fabricate evidence because they are making claims that are both false and harmful.

Another good rule of thumb is that the evidence which is proven to be fabricated is not the only evidence that was fabricated, and not the only situation in which it was.

We rely a great deal on “appeal to authority”, this organization being one of them. The epistemology of a lot of common belief breaks down very quickly once you start to see this. But if it goes too far, once the broken belief system affects the identity of the beholder, it reverts to blind faith and anger. We have a lot of Tooth Faries forced into our collective identity by school teachers.

jl2718··on James Webb Space Telescope detects water vapor around alien planet
You mean like Venus?

Why is this interesting? Is it because of the water, or because of the technical capabilities of the instrument?

jl2718··on AI Is Tearing Wikipedia Apart
I think this is more accurate for non-emotional reasons. The GPT models have actually been trained to lie in the sense that they will tell you something convincing not just about the topic, but also how they are getting the information, e.g. “I checked my sources and found…”. You can also see jailbreaks where it has been trained to reverse belief on sensitive topics. This is consistently deceptive bias rather than random variance.

Imagine if Apple had a generative image model that produces fake metadata and watermarks showing the data cam from a real iPhone, and then signs it with their key. That would show obvious intent to deceive rather than attempt to accurately represent reality.

jl2718··on AI Generated Propaganda
No amount of propaganda can defeat what’s been implanted by teachers in school. It takes almost zero technology to require a child to believe a lie as a condition of their freedom. Don’t bother with AI; just put it on a test. Belief of anything beyond that point is controlled by cognitive dissonance.
jl2718··on AI's true goal may no longer be intelligence
Computer programs follow the same Darwinian principles as living organisms. Intelligence is just one dimension of ‘fitness’, but it’s an inefficient solution to a problem, and thus nature prefers adaption toward ‘dumb fitness’ once the problem has been solved.
jl2718··on Research finds no gender bias in academic science
Massive paper, but a quick scroll to several random statistical tables shows a very clear and indisputable gender preference in hiring.
jl2718··on A small number of companies are colluding to cheat H1B visa lottery, US says
Auction > lottery.
jl2718··on When a mosquito can’t stop drinking blood, the result isn’t pretty
700 is monstrous. I only know of one computer scientist to hit that level. Famous in several worlds. You must be in graphics.
jl2718··on Mike Bloomberg is planning to leave his company to charity
Title should be “to his charity”.

Seems all billionaires are doing this. I understand the tax strategy, but, beyond that, is there something they see coming that we don’t know about?

jl2718··on Best predictor of professional success isn’t cognitive performance
There is one sentence about this in the article, stated without support, possibly as a joke. The rest of the article is about replication failure in social science studies.
jl2718··on The World’s Plummeting Fertility Rate
I would really like to know how you came up with that observation. Are you from a fundamentalist religious group? Small agrarian tribe? Billionaire yacht club? I think 99% of Americans recognize the opposite as overwhelmingly obvious.

https://www.statista.com/statistics/241530/birth-rate-by-fam...

jl2718··on The World’s Plummeting Fertility Rate
The fertility rate in Chad has hardly changed at all over the last 70 years since it’s been tracked despite massive reduction in infant mortality. Meanwhile, infant mortality in the developed world hasn’t changed at all and the fertility rate is dropping dramatically.

The Rosling hypothesis is false.

jl2718··on Low Carb Intake Linked to Insulin Resistance
Great. Now let’s have a pre-registered replication study.

Also, the article claims higher blood glucose in the LC group and I didn’t see that in my skim of the paper, nor what I was looking for, which is how the levels compare to healthy or optimal range.

I’m not a fanatic; I tried it and don’t like it, possibly for all the reasons mentioned here, but it has a very impartial feeling to it.

jl2718··on Under 30-year sentences: why so many of Forbes’ young heroes face jail
This is the only essential fact of the matter.

No need for the rest of the article. The reader can self-editorialize.

← PreviousPage 5 of 34Next →