a good spy cam doesnt have strong RF emitting output. Noway those cheap gadgets will detect those spycams.

When you don't know the answer to a question you ask an LLM, do you verify it or do you trust it?

Like, if it tells you merge sort is better on that particular problem, do you trust it or do you go through an analysis to confirm it really is?

I have a hard time trusting what I don't understand. And even more so if I realize later I've been fooled. Note that it's the same with human though. I think I only trust technical decision I don't understand when I deem the risk of being wrong low enough. Overwise I'll invest in learning and understanding enough to trust the answer.

For all these "open questions" you might have it is better to ask the LLM write a benchmark and actually see the numbers. Why rush, spend 10 minutes, you will have a decision backed by some real feedback from code execution.

But this is just a small part from a much grander testing activity that needs to wrap the LLM code. I think my main job moved to 1. architecting and 2. ensuring the tests are well done.

What you don't test is not reliable yet, looking at code is not testing, it's "vibe-testing" and should be an antipattern, no LGTM for AI code. We should rely on our intuition alone because it is not strict enough, and it makes everything slow - we should not "walk the motorcycle".

Ok. I also have the intuition that more tests and formal specifications can help there.

So far, my biggest issue is, when the code produced is incorrect, with a subtle bug, then I just feel I have wasted time to prompt for something I should have written because now I have to understand it deeply to debug it.

If the test infrastructure is sound, then maybe there is a gain after all even if the code is wrong.

> I have a hard time trusting what I don't understand

Who doesn't? But we have to trust them anyway, otherwise everyone should get a PhD on everything.

Also for people who "has a hard time trusting", they might just give up when encountering things they don't understand. With AI at least there is a path for them to keep digging deeper and actually verify things to whatever level of satisfaction they want.

Sure, but then I rely on an actual expert.

My issue is, LLM fooled me more than a couple of times with stupid but difficult to notice bugs. At that point, I have hard time to trust them (but keep trying with some stuff).

If I asked someone for something and found out several time that the individual is failing, then I'll just stop working with them.

Edit: and to avoid with just anthropomorphizing LLM too much, the moment I notice a tool I use bug to point to losing data for example, I reconsider real hard before I use it again or not.

Often those kind of performance things just don't matter.

Like right now I am working on algorithms for computing heart rate variability and only looking at a 2 minute window with maybe 300 data points at most so whether it is N or N log N or N^2 is beside the point.

When I know I computing the right thing for my application and know I've coded it up correctly and I am feeling some pain about performance that's another story.

I tell it to write a benchmark, and I learn from how it does that.

IME I don't learn by reading or watching, only by wrestling with a problem. ATM, I will only do it if the problem does not feel worth learning about (like jenkinsfile, gradle scripting).

But yes, the bench result will tell something true.

You could try Alan wake. It really gets similar vibes.

Not entirely coincidentally, given that both are Remedy games and share some key creative staff, in particular Sam Lake [1] and Petri Järvilehto.

[1] https://en.wikipedia.org/wiki/Sam_Lake

Syndicate PC game vibes

Indeed. The first Syndicate game is still in my memory due to the UI alone. I don't remember getting very far in it though, too difficult.

I once liked the EU. Well still do it because of the east to travel without borders. But it's leadership is something dangerous and may shape to some form of dictatorship or entity that does not serve its people. But a small minority consisting out of some large companies.

This is a proposal from one wing of polititians that still hasn't even passed a basic voting process in EU parliament.

So what exactly are you screeching about? Which nation on this world has leadership that never proposes anything like this? Which one is 100% pure and noone even thinks about bad things to bring up to a vote?

We should absolutely be “screeching” about any attempt to push or rebrand chat control. You’re framing it as some sort of business as usual type situation where a dumbass politician does a thing. Absolutely not the case and you’re either uninformed or a bad actor. There’s layers of lobbying, corruption and ideology to unpack here https://mullvad.net/en/blog/mullvad-vpn-present-and-then

The EU leadership is the leaders of the 27 sovereign countries

Now you can argue there is a democratic deficit in those countries, sure.

No, the European Commission and the unelected bureaucrats behind it are the true leaders of the EU.

They are the only long lasting institution that can do time arbitrages (wait for the right presidency to push new regulations), they have the means to pressure individual MPs, and they are the ones holding the pen during the negociations between the parliament and the States. The EC is also the master of the legal agenda, the banana republic-style parliament can't decide which laws they vote.

Because the EC has little to no budget to spend, and its only tool is regulation (that doesn't require cost/benefit analysis btw), they...spend their day regulating. They are not constrained by execution either since the States are in charge of applying and dealing the regulations, however how detached from reality they are.

The EC bureaucrats come from a small elite, remote from the reality of the common man. Ursula Von der Leyen is a good example of this. Fun fact, a phd is required to become a EC bureaucrat, so many of them...just buy the services of a post-doc researcher to write it for them. I used to work with a colleague who did it as a side job.

There's democratic deficit in the whole system as this issue wasn't part of most internal election campaigns, effectively circumventing democratic process, due to lack of input from citizens themselves.

EU severely lacks checks and balances if it tries to be something more than trade union.

Are you suggesting the existence democracies that only ever implement policies that were a significantly theme during elections?

The EC has no democratic control, their members are not elected, only the commissioners are "approved" by the european parliament. Its actions are also obfuscated and mostly non-public (as we saw in the ChatControl case, for instance), so citizens hardly even know what's happening.

Dunno about European governemt but us secretaries of state are appointed by the head of government just like a commissioner is

The president is responsible for the secretaries of State actions, and more generally their respective parties.

No one is responsible for the commissioners' actions, and they can't be fired. When Von der Leyen lied and refused to show her text messages where she privately negotiated Covid vaccines, nothing happened. When the EU commissioner for digital markets left and got hired by Uber right after... nothing happened, as no one was responsible.

Commissioners hold the legislative power, as they choose which laws to introduce and hold the pen during negociations. It's pure, unchecked bureaucratic power that ends up with a never ending flow of stupid regulations that weaken Europe slowly.

I'm suggesting that there are enough layers of interdiction, that you can easily 'wash' political fallout and push legislation that would otherwise get you voted out of office in local elections.

In the U.K. you can vote for local councillors all you want, won’t make any difference to Westminster

I think EU will manage without you liking it. But painting its leadership as the one trying to shape dictatorship is incredible ignorant.

Europe is preparing for the Russia invasion from one side, and betrayal by the US from the other.

A country serving small minority of large companies is the best description of the US, not the EU.

Wow. I cannot fathom anyone thinking this, but also I am doubtful the EU pays for propaganda on HN so it is what it is I guess. After von der Leyen's corruption and the fast pace into totalitarianism against the will of the population nonetheless. Just wow.

Are you really convinced that the EU, which is not even a nation and is usually laughed away for being incapable of making any firm decision whatsoever, is on a faster track towards totalitarianism than the US has been since its last election?

No, and I didn't claim that - but it is sliding at a fast pace too.

> Europe is preparing for the Russia invasion from one side, and betrayal by the US from the other.

Let's assume for a moment that would be true. And let's also ignore the lack of a nuclear weapons in most EU countries.

How does breaking encryption for normal people help? Spies and Operatives will just use PGP and ignore these laws, because that's what spies do.

Mind you I don’t believe this, but the logic is if encryption is banned, then anyone using it will be easier to find like spies.

Before online encryption, spies still used code books but having one in your house was essentially proof you were a spy.

Didn’t spies just use common books like war and peace or the bible

I'm as pro european as they come, but I think the author didn't deserve a downvote.

If there is a moment when the EU could not afford to take hits to their popularity, it is now. And here we are, gifting free shots to anti-EU populists.

Measures such going dark and similar ones are wholly supported - and pushed - by police forces around europe, not by politicians. I do agree that the politician should grow a spine and trust computer scientists for one, since they're the ones making laws after all

> I do agree that the politician should grow a spine and trust computer scientists for one

Trust the computer scientists on how to prevent crime? Uh, well that's certainly creative.

No, trust the computer scientists on what can easily be circumvented by criminals while still allowing third parties to scan private conversations. But I do suspect a bit that this is only an intended side effect

As opposed to blindly trusting the police and LEA? Yes, absolutely — I'd rather trust computer scientists.

Your description match the US as well.

APN/Kibana. All what I need for inspecting logs.

Shoutout to Kibana. Absolutely my favorite UI tool for trying to figure out what went wrong (and sometimes, IF anything went wrong in the first place)

I vowed to never book at booking. They use dark patterns and are therefore a parasite. I really wish them to go bankrupt with all the responsible ones going too. I really don't understand they we as a so called caring society accept behavior of these kind of companies.

If you want to try out other models try opencode. Right now grok is free to use. I am using it now. I think its a little better than codex or Claude. But it's so so much faster. Gemini 3 can also be used, but is often overloaded.

Why is it that all these languages like PHP, but also typescript are becoming like impossible puzzles to read. I find these generics, types and other language features very often causing complex software architecture. I see so many collegues these days struggling in understanding codebases. You almost need a PHD brain to be a frontend web developer.

Counterpoint (from the same website): https://stitcher.io/blog/evolution-of-a-php-object

PHP 8.2 has this:

``` readonly class BlogData

{

    public function __construct(

        public string $title,

        public State $state,

        public ?DateTimeImmutable $publishedAt = null,

    ) {}

}

```

Whereas in php 5.6, to accomplish the same you need all this:

``` class BlogData { /* @var string / private $title;

    /** @var State */
    private $state;
    
    /** @var \DateTimeImmutable|null */
    private $publishedAt;
   
   /**
    * @param string $title 
    * @param State $state 
    * @param \DateTimeImmutable|null $publishedAt 
    */
    public function __construct(
        $title,
        $state,
        $publishedAt = null
    ) {
        $this->title = $title;
        $this->state = $state;
        $this->publishedAt = $publishedAt;
    }
    
    /**
     * @return string 
     */
    public function getTitle()
    {
        return $this->title;    
    }
    
    /**
     * @return State 
     */
    public function getState() 
    {
        return $this->state;    
    }
    
    /**
     * @return \DateTimeImmutable|null 
     */
    public function getPublishedAt() 
    {
        return $this->publishedAt;    
    }

} ```

I assume it is some inferiority complex, on many sides. PHP itself was laughed at being too simple, underpowered and inconsistent, now they overcorrected with types, annotations and breaking backwards compatibility with every release so that no old code base can remain intact. Frontend devs yearned to be regarded as real developers, which in their context means construction of unwieldly and overcomplex enterprise bullshit, thus typescript etc. And in the backend you have that same mechanism, devs having to prove they are no beginners and thus using (wrongly) design patterns, instructed by software architects, instead of avoiding abstraction and thus complexity.

No, I'm not bitter.

Can you share an example of what you're talking about in PHP 8.5? On the linked web page, the only code pattern that looks remotely complicated to me is the following:

    #[SkipDiscovery(static function (Container $container): bool {
        return ! $container->get(Application::class) instanceof ConsoleApplication;
    })]
    final class BlogPostEventHandlers
    { /\* … \*/ }

It's somewhat comforting to read my insecurities shared by others in this thread

Which parts of PHP have become harder to read?

Like on of my sibling comment, I truly believe this is connected to some degree of social pressure.

People pointing fingers to "outdated" languages for not having some of the most trendy constructs.

The pipe operator is definitely one of the feature that create more ways to do the same thing while providing unclear benefit.

Never in my life I was in a situation like "with the pipe operator this I would have saved me hours of debugging/reading/creating code".

There will be always lack of supply of housing. Its such an essential thing that it never becomes unused or unwanted. The combination of bad regulations where banks can give extremely high loans is the biggest issue. I would not be surprised if mortgages of 60 years will be possible where your total down payment is 4 times the actually price of the house.

Not true. As an overly simple thought experiment consider what happens to the market if a government were to dictate the construction of two units of housing per capita.

The issue is lack of supply. That is enforced by zoning, which is ultimately dictated by politicians, who know they won't get reelected if they make policy changes that gut a significant portion of their constituents' savings.

It's a fairly straightforward example of systemic dysfunction with no obvious solution.

Plenty of places in the world with an oversupply. Lots of examples in China

China (like the USA) didn't let the market handle it. China built without letting market demand dictate supply. USA did the opposite, didn't builders meet supply. They try, they try all the time to built more, they get stopped by nimbys and over regulation.