HNHacker News
TopNewBestAskShowJobs

richk449

815 karma · joined December 11, 2016

submissionscomments
richk449··on AI-Generated Data Can Poison Future AI Models
In many areas of communication and information, this exact problem is dealt with through error correction codes. Do AI models have built in ECC?
richk449··on AI-Generated Data Can Poison Future AI Models
Kessler syndrome for the internet?
richk449··on Where is Noether's principle in machine learning?
Sounds like some of the trendy analogy computing approaches, like this one for example:

https://www.microsoft.com/en-us/research/uploads/prod/2023/0...

richk449··on "Dune" and the delicate art of making fictional languages
My favorite kind of sci fi is the stuff that throws you into a new world with new cultural references and doesn’t bother to explain them. Gibson is a good example. It’s a similar good feeling to getting off the train in a foreign city pre smartphone.
richk449··on Detroit's abandoned tunnel systems open door to another world
Interesting article, but surprisingly poor writing for a flagship big city newspaper. It’s full of typos and incomplete sentences. Is it AI generated text, or just a frantic writer with a dozen more articles to crank out today?
richk449··on Ask HN: What's the best way to make a RC car drive around a permitter?
Donkey car works really well. It uses a camera and by default trains it to follow a lane line. I bet it could be trained to follow an edge pretty easily though.

It’s a well documented and easy to implement project.

https://docs.donkeycar.com/

richk449··on Moving Fast and Breaking Things Is the Opposite of Tech Optimism
The essay seems well reasoned, but the one example he gives looks misguided.

> But… that gets everything backwards. The reason for this excessive caution around nuclear power was the lack of thoughtful and careful early deployments, leading to disasters like Three Mile Island and Chernobyl. It’s perfectly reasonable to suggest that the precautionary principle has gone too far, and overreacted to such a degree that we’re holding back useful nuclear deployments, but if we had been more careful and thoughtful in the early deployments of nuclear power, such that meltdowns were not something we had to deal with and the risk was effectively zero, then we’d see much more nuclear power around the globe.

Three mile island was about as good of an example as you will get of well designed carefully considered deployment of nuclear power. Yes, mistakes were made, and the core melted, but because the plant was generally well designed, no significant contamination of the environment occurred, and no danger was posed to the public.

richk449··on The surprising connection between after-hours work and decreased productivity
I thought your comment was spot on. I just appreciated the possible turn towards nihilism or existentialism.

We spend so much of our life determining what is useful in one sense of the word, and rarely stop to think about the usefulness of what we do in the philosophical sense.

richk449··on The surprising connection between after-hours work and decreased productivity
> And this could be extrapolated to saying everything is useless.

Now we are getting somewhere.

richk449··on After OpenAI's blowup, it seems pretty clear that 'AI safety' isn't a real thing
> But the fear mongering is a bit premature.

I’m sure you would also accept that at some point it could become too late to start worrying about the risks if the danger has already gone exponential.

So how do you determine when is the right time to worry? If you have to error on worrying too early or two late, which way do you bias your decision?

richk449··on Before Altman’s ouster, OpenAI’s board was divided and feuding
Is there a name for this type of analysis?
richk449··on Panama Canal is so congested that one ship owner paid $4M to skip the line
I see. Thank you for the interesting article.

The Tragedy of the Commons as a historical event in Medieval England has been criticized (hard to say debunked from this one article, but perhaps there are more).

The Tragedy of the Commons as an economical/social phenomena has not been debunked.

Since 99.9% of the uses of the phrase Tragedy of the Commons are referring to the second case, I'm not sure it is helpful to say that it has been debunked.

richk449··on Panama Canal is so congested that one ship owner paid $4M to skip the line
Can you tell me more about how the tragedy of the commons is a myth? I am no economist, but I thought it was a straightforward and well understood outcome.
richk449··on How to boss without being bossy
Devils advocate: maybe your boss wanted the team to function autonomously, and was avoiding making decisions that could be made at your level?
richk449··on Can you use your "free will"? Try your hand
It’s explained on the linked site that the free will comment is a joke.
richk449··on Magnetically Modelling MagSafe
Any idea where I could buy this? It would be really helpful for my work.
richk449··on Working in Silicon Valley was fun. Now it's just another miserable corporate gig
Why is that?
richk449··on AI beats human sleuth at finding problematic images in research papers
Yea, that’s super annoying!

Just like the word car doesn’t mean anything. Last time someone threw out the buzzword car to me, it turned out they were actually talking about a 2008 four door Honda civic with four speed automatic in blue paint. I don’t know why they couldn’t have said that instead of trying to bs me by saying they were buckling their kid in the booster seat in their car.

richk449··on What are emerging/promising areas of tech to get into today?
Anything revolutionary you are aware of? Knowing that we will need one soon, and dreading the usual options, I’m ready for a revolution.
richk449··on What are emerging/promising areas of tech to get into today?
Seems like most true technology breakthroughs come from materials science. It is much harder to get the benefits of those advances though, since the path from materials science to product can be long and complicated.
richk449··on Da Vinci robot surgeon removes inoperable tumor, saving patient’s life
Thanks for the informative reply. Your comment rings true to me, since I am aware of other medical systems that are marketing tools.

I have questions though.

Is the davinci amazing? I have heard that it is.

Why doesn’t someone else make a similar product? Is there technology that is hard to replicate?

richk449··on Da Vinci robot surgeon removes inoperable tumor, saving patient’s life
Yea, amazing, not perfect.

If DaVinci hadn’t developed the robot, would more people be able to afford health care?

richk449··on Da Vinci robot surgeon removes inoperable tumor, saving patient’s life
Anyone know more about robot assist systems in general? Are there similar systems for performing delicate assembly operations on a production line? (Non medical)

The die binders that I used to use had physical motion reduction systems - are there more general versions of these available?

richk449··on Da Vinci robot surgeon removes inoperable tumor, saving patient’s life
What an amazing world we live in. There is nothing interesting about a robotic assisted surgery able to perform more delicate and sophisticated operations than the best human surgeon in the world.
richk449··on Rage Against the Machine Used Capitalism to Sell Communism
Great article.

It seems like the classic music industry is in many ways like the VC model. Fund/invest in many startup bands, with the goal that one in a hundred get huge and return massive profits.

The end of the article talks about Bandcamp as an alternative, which is in some ways analogous to YC.

Is this an acknowledged similarity across industries? Are there other industries that have a similar model?

richk449··on All of Physics in 9 Lines
I assume you are joking, but for others reference, this is a great way to answer that question:

http://ppc.inr.ac.ru/uploads/476_Hamill.pdf

richk449··on Ask HN: Any interesting books you have read lately?
Aside from semantics, that doesn’t seem to contradict Sapiens:

In areas such as the Great Plains of North America, where not a single wheat stalk grew 10,000 years ago, you can today walk for hundreds upon hundreds of kilometers without encountering any other plant. Worldwide, wheat covers about 2.25 million square kilometers of the globe’s surface, almost ten times the size of Britain. How did this grass turn from insignificant to ubiquitous?

Wheat did it by manipulating Homo sapiens to its advantage. This ape had been living a fairly comfortable life hunting and gathering until about 10,000 years ago, but then began to invest more and more effort in cultivating wheat. Within a couple of millennia, humans in many parts of the world were doing little from dawn to dusk other than taking care of wheat plants. It wasn’t easy.*

richk449··on Cargo Cult AI
I don’t think so.

Can you tell the shape of a vase by looking at it?

Would you say “yes” or would you infer that the speaker was asking if you could distinguish all unique vases from one another by looking at them?

richk449··on Cargo Cult AI
The author uses the example of Kepler and Newton to claim that NN based AI can’t generalize like humans can, so it can’t become a AGI.

Was Urek, the first caveman able to generalize a physical theory of orbital motion when he looked up at the sky? It took humans millions of tries and hundreds of thousands of generations to come up with Kepler. Why does he give up on NNs so quickly?

richk449··on Cargo Cult AI
“I have recently taken to asking candidates who interview with me for research positions whether it is possible to hear the shape of a drum. This seemingly innocuous problem was posed by mathematician Mark Kac in 196612 and stumped the mathematical community for several decades.

The quick answer I am often given is, "Yes, of course hearing a drum's shape is possible. All that is needed is a sufficiently large dataset of sounds associated with drumhead shapes (for supervised learning) or indeed even without association to the shapes (for unsupervised learning) and use of an effective training algorithm and validation methodology. Once a model has been trained on the data, it will infer the shape of a drum from any recorded spectrum it is given."

This answer is wrong, and it is the reason Kac's famous question merits being revisited in the context of today's AI to solve complex problems. In the 1990s, mathematicians finally proved that it is, in fact, not possible to hear the shape of a drum, or at least not uniquely.9 This is because drumheads exist of different shapes that produce exactly the same sound, or in mathematical terms, are isospectral.”

That seems entirely unfair. The question is if you hear the shape of a drum. The actual answer is yes. But he says this answer is wrong, because you can’t uniquely hear the shape of the drum.

I’m trying to decide if this goalpost moving is a better metaphor for this full piece, or for much of the analysis of AI in general.

← PreviousPage 3 of 14Next →