Hacker Newsnew | past | comments | ask | show | jobs | submit | samrus's commentslogin

They really shouldnt have read all the changes individually. What you gotta do is set up your VC properly so these changes are seperated from good code, and then review the whole set of changes in an IDE that highlights them, like a proto PR. Thats far far less taxing since you get the whole picture

"You know whats alot harder? The first 50%" - windows

Given how Windows 11 is going I would guess "You know what's a lot harder? Not trending downwards." would be the more obvious one.

I heard this in a Justin timberlake as Napster guy voice

This is far too negative and reductionist

Like saying theres no interesting discussions about programming. Just whether OOP is overhyped, python is slow, how well you can convert a c codebase to rust


Trump's stupidity is why italians are suffering from this oil crisis. It makes sense for them to not want to support it in anyway

[flagged]


> If the US had pulled out as Iran demanded and said "not our problem" you'd now have an Iran full-on supporting Russia, providing missiles, drones, and more, control over 20% of the world's oil supply where they can exact tolls, turn Gulf States into vassals (for those who can't get a nuke), and would continue to directly pay, arm, and supply groups throughout the Middle East and elsewhere to destabilize other countries and kill people including in Israel, Lebanon, Yemen, and more.

You're describing the current situation. The tolls are new, as a response to the war! They've been supplying Russia with drones for years! They've been funding groups throughout the Middle East for decades!


[flagged]


> They very well may have been planning to do so in the next 5 or so years.

Our accelerating that doesn't seem like a big win so far.

> They could at a whim, just say "give us X or we close the straight".

They've been able to do that for decades.

> but maybe Iran shouldn't be doing these things

Few disagree.

But you're falling for https://en.wikipedia.org/wiki/Politician%27s_syllogism.


> Our accelerating that doesn't seem like a big win so far.

Maybe this can help.

It's 2025 and Iran has 1,000 missiles (just random rough numbers).

It's 2030, and Iran now has 5,000 missiles and bought a bunch of hypersonic missiles from China.

Sure, we can just "not accelerate" this, but then we're not going to do anything about it in 2030 because nobody is going to accept being on the receiving end of so much destruction.

> They've been able to do that for decades.

Not quite - they've had varying levels of capability here. Until recently they haven't had the missile and drone stockpile that they have today and have been building.


its not either or. if we expected support from European countries, we could have given them a heads up and understood their position in advance. it was not an emergency situation. we had an agreement, I know that doesn't mean shit to the current administration, but not every country is willing to go cowboy, and give up their sovereignty just because the US doesn't like rules and laws anymore. the US would clearly be in a better position now if it hadn't said 'screw you guys' and started shooting.

I largely don't disagree with you, but at the same time had Trump asked they would have told him no anyway because Europe has little to no ability to help here and they're scared. As European prime ministers and officials have said already, and I'm paraphrasing "what can a few frigates do that the mighty US Navy cannot?". It rings true, as did the comment from the I believe Polish prime minister (or perhaps the foreign minister, either way) which said 600+ million Europeans are asking 300+ million Americans to defend them against 180+ million Russians. Something doesn't quite add up here.

I mean, they are controlling Hormuz Strait now - they were not before the war. In addition they are now free to sell their own oil, because the sanctions have been lifted. And I can't see why they would stop supporting Russia with drones either. They also got to keep their enriched Uranium. I don't think I can name a single reasonable objective that US have achieved with this war.

> And I can't see why they would stop supporting Russia with drones either. They also got to keep their enriched Uranium. I don't think I can name a single reasonable objective that US have achieved with this war.

It's been 3 weeks... we're going to stop the enrichment. We can stop the drone supply by blowing up the factories.

> I mean, they are controlling Hormuz Strait now - they were not before the war. In addition they are now free to sell their own oil, because the sanctions have been lifted.

They were de facto "controlling" the straight anyway because of their missile stockpile, they just hadn't yet exercised their leverage. They're free to sell their oil to the extent the US allows them to based on whatever decisions are being made. Obviously we can just blockade the straight too, even more effectively than Iran has been.


This is a brilliant example of DARVO: The US and Israel decided to attack Iran but somehow they had no choice and are actually the victims in all of this.

What about cops and legislators? They thing AI is infallible and thats very convenient for them since they can thus not mandate cops having to double check tmwhat the AI suggests

Flushing context is something i've found to be critical. If the cintect gets poisoned then you gotta start a new chat or it wont recover.

And you need to migrate the validated context from the old chat to the new one. For that it helps to hage it document the established and validated points as checkpoints.


Glib, but it wont be cost effective at that small scale

So are we arguing that the article that talks about them using ASICs is just making that up then? Otherwise what's the fourth option?

Who says CERN needs to be cost effective?


Thats a bit far. Relu does check x>0 but thats just one non-linearity in the linear/non-linear sandwich that makes up universal function approximator theorem. Its more conplex than just x>0

Multiply-accumulate, then clamp negative values to zero. Every even-numbered variable is a weighted sum plus a bias (an affine transformation), and every odd-numbered variable is the ReLU gate (max(0, x)). Layer 2 feeds on the ReLU outputs of layer 1, and the final output is a plain linear combination of the last ReLU outputs

    // inputs: u, v
    // --- hidden layer 1 (3 neurons) ---
    let v0  = 0.616*u + 0.291*v - 0.135
    let v1  = if 0 > v0 then 0 else v0
    let v2  = -0.482*u + 0.735*v + 0.044
    let v3  = if 0 > v2 then 0 else v2
    let v4  = 0.261*u - 0.553*v + 0.310
    let v5  = if 0 > v4 then 0 else v4
    // --- hidden layer 2 (2 neurons) ---
    let v6  = 0.410*v1 - 0.378*v3 + 0.528*v5 + 0.091
    let v7  = if 0 > v6 then 0 else v6
    let v8  = -0.194*v1 + 0.617*v3 - 0.291*v5 - 0.058
    let v9  = if 0 > v8 then 0 else v8
    // --- output layer (binary classification) ---
    let v10 = 0.739*v7 - 0.415*v9 + 0.022
    // sigmoid squashing v10 into the range (0, 1)
    let out = 1 / (1 + exp(-v10))

i let v0 = 0.616u + 0.291v - 0.135 let v1 = if 0 > v0 then 0 else v0

is there something 'less good' about:

    let v1  = if v0 < 0 then 0 else v0 
Am I the only one who stutter-parses "0 > value" vs my counterexample?

Is Yoda condition somehow better?

Shouldn't we write: Let v1 = max 0 v0


The relu/if-then-else is in fact centrally important as it enables computations with complex control flow (or more exactly, conditional signal flow or gating) schemes (particularly as you add more layers).

I feel like this is a somewhat bad faith argument becuase its shifting blame onto the user.

Its addicting the same way gambling is addicting. Its ridiculous to say that people who get addicted to it are not neurotypical.


Yeah. I think we should let the free market decide if our children do heroin too. Whatever happens, at least it'll be economically efficient

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: