HN user

joantune

153 karma

[ my public key: https://keybase.io/jantunes; my proof: https://keybase.io/jantunes/sigs/5HV4DEP2e-cs0w65p8Er411re1yiU7VGHt-nkqHObv0 ]

Posts6
Comments181
View on HN

But, all we’ve achieved at this point is making a glorified token predicting machine trained on existing data (made by humans), not really being able to be creative outside of deriving things humans have already made before. Granted, they‘re really good at doing that, but not much else.

Remove token, and that's what we humans do.

Like, you need to realize that neural networks came to be because someone had the idea to mimic our brains' functionality, and see where that lead to.

Many skeptics at the beginning like you discredited the inventor, but he was proved wrong. LLMs shown how much more than your limited description they can achieve.

We mimicked birds with airplanes, and we can outdo them. It's actually in my view very short sighted to say we can't just mimic brains and outdo them. We're there. ChatGPT is the initial little planes that flew close to the ground and barely stayed up

The intensity you did it though matters. You probably didn't spend that many years on a specific sport for instance.

And when we're talking about sports, genetics matter as well (depending on each one)

When we're talking brains, while genetics also matter, assuming normal (whatever that is) brain, the plasticity changes a lot how it operates.

So, the 10 years thing is definitely a big if not the biggest part. In my opinion. Would love to see studies if any exist out there on this

They're not boasting about their current productivity, they're boasting about the one they achieved at FoundationDB when they implemented the testing, which gave them the idea to build antithesis

Perhaps just give descriptive names and then change it when the scope changes??

"Problem" solved!

Afraid of losing stars/dlds on the repo/etc? Make it clear in the Readme where the new project is. Have warnings et al for NPM projects.

Ask npm/GitHub to allow name changes with redirects.

This proposed "solution" is just bad for many reasons:

You need to delve into the Readme to find out what it does. Imagine you're seeing a package.json and for each line you have to Google what's the project about

SEO is best if name matches.

These are enough *good* reasons why this advice is a bad idea

I see zero examples and rational/practical arguments against the usage of REST.

The only thing here is: it wasn't intended to be used this way.

I see it being called an overload, but I see no alternatives..

This article adds little of value, or even doesn't highlight or enunciate the disadvantages of using REST for everything.

And on the other hand, the advantages are obvious

Show HN: Fraidycat 6 years ago

Circa 4 weeks ago I had the same core idea! I think this is very interesting to be able to follow people you respect or are interested in one way or another, regardless of the platform where you find them. A kind of global Twitter. This has potential to become the new version of a social network. Kudos on executing it

PS: Initially I thought that the killer platform to adopt such a thing would be keybase! I even thought of tweeting to them, but then the idea just fleeted away. Oh well..

Yes! So, from my experience it's invaluable to: See why things were done like that. My best usage so far of it was to add a #number at the end of the commit message where number is a trello card id. Then you can go back to the discussion and card that originated that change. You can see all the context of why something was done like that.

This becomes very valuable when maintaining projects that will be running for years, and prevents you from undoing things or going back to doing the same mistakes

That's why TrueCrypt, and I think Vera Crypt have support for a volume that can be decripted with two keys - one will yield a decoy volume where its free space is actually the real volume that you want to protect which gets decoded only with a second key

I'm not sure if the entropy analysis of that free space can suggest that there's something funky about that free space or not.. Because usually free space is either actual info just marked as deleted, or info reset to zeros by some pro active wiping of the free space. So, having a bunch of whacky data that doesn't look like any kind of file, can probably be used as a tell tale sign? No?

This is awesome! I would hook this up with Shopify and one of those printing and shipping services and see where that leads you. I'm a Shopify expert, already did that for customers, if you need advice let me know

Well, in RSA you have to choose two prime numbers, multiply them, and keep them secret. p and q: pq =n . And n is made public. I wonder if this probability, maybe coupled with concrete implementations, makes it for a more restricted set of guessing p and q. That would be it. I guess that given that this only introduces 'restrictions' on sequential prime numbers, doesn't really help at all, given that p and q should be random. Unless there's a shortcut applied in implementation that you find a random p and then q is the next prime number. Hence my question to the community. But I only know that both RSA and DH rely on prime numbers.

I wonder what the implications are for cryptography - given the heavy reliance on 'random' prime numbers

but one would have to delve into this - and from the top of my head there are no algorithms that make you pick 'two consecutive' primes - so the implications - if any - aren't obvious - anyone here has some more light or other angles to shed into this?

Google Maps Hacks 6 years ago

From my last experience circa 1/2 years ago, you just need to be the only one on the road. About that time I was driving up at odd hours on a highway, when I noticed a yellow stretch on google maps coming up. Turns out the highway was mostly empty, and the only other vehicle in the road was a truck, which is limited to a fraction of the top speed.

How to Exit Vim 7 years ago

I can't wait for this article to be on top of Google's search results for 'how to exit vim' = D

Nice! I was just looking to convey entity diagrams to a client - and thought of doing them through something like Omnigraffle - this looks like it could do the job.

However - it doesn't seem to export itself well (the Print doesn't work - or something like a read only link would be enough)

Those are quite complex and evolved processes to learn in the first place to start with. It must be something more than trial and error, right? Am I the only one thinking that if a plant killed me, or slowly made my legs paralyzed, more likely one would abandon that process rather than say: let's try to roast it and/or leave it outside for a couple of days. Unless they were desperate and had no other food source.

The case with the cake is even more mesmerizing, it slowly blocked overtime their ability to process B1. How could people tell that it was from that food?? And how do they go, yeah, we just forgot to do X