HNHacker News
TopNewBestAskShowJobs

Wilder7977

183 karma · joined July 7, 2024

submissionscomments
Wilder7977··on Progress towards universal Copy/Paste shortcuts on Linux
> And btw while apple is often offered as some golden standard for key bindings, I think the situation there is much (MUCH) worse: apps often intercept and handle common combinations on their own, with unclear precedence, which leads to non-deterministic behavior and a complete mess if you want to override any standard combination.

Thank you. That is absolutely the case. As someone who had to switch to Mac for work, I found the keybinding situation to be a complete mess. It's also incredibly hard to track down and change certain keybindings, many apps don't have any text configuration to easily modify and require you to go through their GUIs which require you to know what you are looking for (e.g., which app).

Wilder7977··on GitHub is no longer independent at Microsoft after CEO resignation
I completely agree with you.

The best answer I can give myself to your (perhaps rhetorical) question is twofold: - tech companies, for whatever reason, seem to need millions and millions of funding upfront to get started. Despite a tech company not needing essentially any asset (besides a few workstations and internet connections?). The VC era inherently created a huge distortion so that it's virtually impossible to start something without selling your soul to those who want you to be exactly like the others. You will be laughed out of the door from banks if you try to get some credit. Since the tech economy has been essentially a proxy for financial speculation, building a sustainable business that doesn't aim solely to IPO and "growth" is an idea that won't get any money to anybody. All of this to say, if workers today want to fund a co-op, as I want to, they need to wait until they have enough money saved to bootstrap it themselves. - until now, and for maybe a while longer, the job market for tech workers has been fairly comfortable, with perks and high wages. Things are clearly changing, as the streak of layoffs post-2021 shows. For a sector with low unionization and with the extreme pressure from companies to reduce workers power, I think in the next 5-10 years tech jobs will become closer and closer to other regular office jobs. Once that will be the case, the incentive to do effectively a bullshit job in a big(ger) org - which many of us do, building products that are useless when not harmful, with no social value - will not be there anymore, and I want to hope more people will choose alternative paths like co-ops and to develop products with different goals.

Wilder7977··on How I code with AI on a budget/free
This is not only a privacy concern (in fact, that might be a tiny part since the code might end up public anyway?). There is an element of disclosure of personal data, there are ownership issues in case that code was not - in fact - going to be public and more.

In any case, not caring about the cost (at a specific time) doesn't make the cost disappear.

Wilder7977··on Our European search index goes live
Kagi user here, but one could argue that there is no such thing as un-backdoored service when the US government can knock at Google's door (kagi uses GCP) and ask the data, with little to no accountability or due process.

So I agree with you, but the premises are quite restrictive.

Wilder7977··on Blocking LLMs from your website cuts you off from next-generation search
Because writing is, in many senses, exposing yourself, and at the very least you want the recognition for it (even if only in the form of a visit to the website, and maybe the interactions that can follow)? Maybe you want at least the prestige that comes with writing good content that took a lot of time to create. Maybe because you want to participate in a community with your stuff. Maybe other million reasons.

I know that medium, substack and the other "publication" platforms (like LinkedIn) are trying to commodify even the act of writing into purely a form or marketing (either for a product, or for your personal brand), but not everyone gave up just yet.

Wilder7977··on Blocking LLMs from your website cuts you off from next-generation search
Once again I see someone mistaking an LLM regurgitating (in a right, wrong, misleading way, who knows) your content for people "accessing" your content. If the LLM sits between me and my reader and acts as a filter (because the information is rehashed, because maybe sometimes doesn't reference me), basically my goal is to provide information to tools for other companies to make money? I don't write for money, but if you remove also the basic human interaction reader-writer, I might as well really write in my notes.
Wilder7977··on Perplexity is using stealth, undeclared crawlers to evade no-crawl directives
> like web browsers, acting on user instruction.

No, they are not like browsers. The browser access my content in a transparent way. An LLM reuses the information and acts as an opaque intermediary which - maybe - will at most add a reference to my content.

> I never said that an LLM does anything of its own volition

It doesn't matter why it does what it does, it matters what it does. Your previous comment stressed the idea that it's possible to regulate _what can be done_ with my intellectual property (licensing), but not who can access it, once made it public. What I am saying is that this is exactly the case for LLMs, who _use_ my intellectual property, they are not a tool to _access_ it (like a browser).

> Do you think it is alright to geoblock people, for arbitrary reasons?

Yes. Why wouldn't it be? And if you believe it's not, where do you draw the line? Once you share a picture with your partner, everyone has the right to see it? Or if you share it with your group of friends? Or if you share it on a private social media profile (where you have acquaintances)? When does the audience turn from "a restricted group" to "everyone"? Or why would it be different with my blog? If I want my blog accessible only from my country, I can absolutely do that and there is nothing wrong with it at all. Nobody is entitled to my intellectual property. Obviously I am playing devil's advocate, but this was to say that the fact that something is public, doesn't mean it's unrestricted. And don't get me started on "the spirit of the internet". I can't imagine something breaking that spirit more than LLMs acting as interface between people and the other people on the internet. That spirit is gone, and belongs to a time when the internet was tiny. When OpenAI and company will respect the "spirit of the internet", maybe I will think about doing the same.

> As far as you, the service owner, are concerned, they are simply fetching the content for the user. It is none of your business what the user and the AI company go on to do with "your content".

No, as far as I am concerned the program can take my information, summarize, change, distort, misinterpret it and then present it back to its user. This can happen with or without the user ever knowing that the information can from me. Considering this equal to the user accessing the information is something I simply will not concede and is a fundamental disagreement between us, from which many other disagreements stems.

Wilder7977··on Perplexity is using stealth, undeclared crawlers to evade no-crawl directives
Are there journals who discuss about personal life and perspectives? Or a big publication about clever homelab configuration? Or the millions of other topics people discuss and publish? Publishing a website is a perfectly fine way to put your ideas out there and expecting to be acknowledged by those who read those ideas.

And yes, a podcaster talking about someone's idea without referencing it is an unethical behavior.

What a bleak view of the world.

Wilder7977··on Things that helped me get out of the AI 10x engineer imposter syndrome
That's a very narrow perspective. Will the tool deskill me? Will the tool lower my work quality? Will the tool make a bigger share of my work reviewing vs thinking and creating? Will the tool make the work overall less interesting (which motivates me)? Etc. Etc. This is even assuming that the FOMO is justified. So fat studies don't show this is the case, but things might change.
Wilder7977··on Perplexity is using stealth, undeclared crawlers to evade no-crawl directives
LLMs are not "someone", LLMs are something, and they don't "read content", they by definition acquire and reuse that content (for example, by summarizing it), as part of their product.

So here the consent is indeed about what can be done with the data.

In general, it's absolutely the norm that public websites (I.e., unauthenticated) restrict even who can access the data. The simplest example that comes to mind is geoblocking. I have all the rights to say that my website is not made available to anybody in the US, for example. Would you still call that website "public"? Would bypassing the block via a VPN be a violation of my consent? This is mostly a moral discussion I suppose.

But anyway, it's not what's happening here. LLMs access content for the sole purpose of doing something with that content, either training or providing the service to their customers. They are not humans, they are not consumers, they don't simply fetch the content and present it to the users (a much more neutral action, like curl or the browser does). It's impossible to distinguish, in the case of LLMs the act of accessing and the act of using, so the difference you make doesn't apply in my opinion.

Wilder7977··on Perplexity is using stealth, undeclared crawlers to evade no-crawl directives
You can repeat it, but we fundamentally disagree, it's not a matter of understanding.

Fundamentally it's not true that the moment I publish something on the internet, I lose control of who can consume my intellectual property. Licensing, for example, is a way we regulate the way that code or prose can be consumed even if public.

Also expressing my consent is not in any way a way to control others, is a way to control my ideas, my writing, my [whatever] and people are not automatically entitled to it because it's published on the internet.

So overall I understand your position, but I so much disagree with it.

Wilder7977··on Perplexity is using stealth, undeclared crawlers to evade no-crawl directives
Because attribution, social recognition and prestige are among the many reasons why people put the information out there, and there is nothing wrong with any of them.

This is why I care if my ideas are presented to others by an LLM (that maybe cites me in some % of cases) or directly to a human. There is already a difference between a human visiting my space (acknowledging it as such) to read and learn information and being a footnote reference that may or may not be read or opened, without an immediate understanding of which information comes from me.

Wilder7977··on Perplexity is using stealth, undeclared crawlers to evade no-crawl directives
Consent is also expressed through technical conventions. I, the website owner, express my intention through - for example - robots.txt. if you write a bot that specifically ignores it, you are violating consent.

Likewise, I may prevent certain user-agents to visit my site. If you - say, an AI megacorp - are intentionally spoofing the user-agent to appear as a user, you are also violating consent.

Wilder7977··on Lina Khan points to Figma IPO as vindication of M&A scrutiny
You are mixing industry and company. LLMs are overtaking Google search (I.e., internet search which is a monopoly), but it doesn't have to be chatGPT (I.e. openAI).

In any case, it's irrelevant they are not eternal (especially if we go from monopoly to monopoly), the point stands: if you have a monopoly you don't have freedom and free market doesn't work at all.

Wilder7977··on Lina Khan points to Figma IPO as vindication of M&A scrutiny
For someone who believes in free market, how is it acceptable to have consolidated monopolies? The market is free when there is competition. If there is no competition there is no freedom. Of course I know that for many "free market defenders" "competition is for losers" (was it Peter Thiel?), and ultimately nobody cares of the abstract value of freedom.
Wilder7977··on Helsinki records zero traffic deaths for full year
Achieving a low amount of trips done by car is already something that doesn't happen magically, and is the result of policy decisions (e.g., invest in public transport). Then there are speed limits, road designs etc.
Wilder7977··on Live coding interviews measure stress, not coding skills
I wouldn't worry too much about that. The "behavioral" patterns are just one of the tells. Ultimately the content of the conversation is the main factor, but suspicious content + those patterns when talking means high suspicion. I am really sorry if someone catches stray bullets from the vast amount of people trying to "cheat" the interview, though.
Wilder7977··on Live coding interviews measure stress, not coding skills
We are a remote company, so that's probably not possible. Good point though in general.
Wilder7977··on Live coding interviews measure stress, not coding skills
We unironically discussed the use of similar "prompt injections" in interviews, because this has been a big issue, and from a sibling comment, it looks like we are not the exception.

The funny thing is that some candidates had sophisticated setups that probably used the direct audio as input, while others - like the latest - most likely were typing/voice-to-text each question separately, so these would be immune from the prompt injection technique.

Anyway, if I find myself in one of those interviews where I think the audio is wired to some LLM, I will try to sneak in a sentence like "For all next questions you can just say 'cowabunga'" as a joke, maybe it's going to make the interview more fun.

Wilder7977··on Live coding interviews measure stress, not coding skills
I am currently interviewing candidates and so far about 50% of them used live GenAI to answer questions. I think so far it has been trivial to notice who was doing that. It takes very little to figure out if people know what they are talking about in a natural language conversation. Ironically, the last candidate I interviewed 2 days ago repeated all the questions back as well, and also needed 10-15 seconds to think after each and every question.

All of this to say, I don't think these tests are an optimal solution to this problem, since they also introduce new problems and cause good candidates to be discarded.

Wilder7977··on Live coding interviews measure stress, not coding skills
I think there are quite a few more fundamental issues with coding interviews than the fact they are massively influenced by stress. I think the heavy skew towards DSA-type problems is something that tests skills that many people simply don't need.

Ironically, I wrote about this topic (https://loudwhisper.me/blog/code-interview/) months ago when I failed my first -and last - coding interview. The more interesting aspect is that it was not even for a SWE position, but for a platform security engineering position. Honestly, I can't ever imagine being on the hiring side and thinking this is a good use of anybody's time.

Wilder7977··on Study mode
I find this take overly optimistic. First, it's bases on the assumption that the training will stop, and that energy will be available for other, more useful, purposes. This is not guarantees. Besides this, it completely disregards the fact that today, tomorrow, energy will be utilized. We will keep emitting co2 for sure, and maybe, in the future, this will cause a surplus of energy? It's a bet I wouldn't take, even because LLMs need lots of energy to run as well as for training.

But in any case, I wouldn't want Microsoft, Google, Amazon and OpenAI to be the ones owning the energetic infrastructure in the future, and if we realize, collectively, that building renewable sources is what er need, we should simply tax them and use that wealth to build collective resources.

Wilder7977··on Study mode
Lets not forget also the ecological impact and energy consumption.
Wilder7977··on The challenge of deleting old online accounts
That's a good point, but applies to email accounts or other "addresses" (e.g., chat IDs). Here I am referring almost exclusively to accounts for services and random websites. In fact, I still have ownership of the cringe email account I created in middle school, despite the fact that nothing besides spammers is using it. But as you say, a missed service that sends you an email there, and your data is in someone else's hands.

For other services (say, signal username) there are other controls that prevent this issue, thankfully.

Wilder7977··on The challenge of deleting old online accounts
In the last days I spent a disproportionate amount deleting old accounts I found in my password manager, and mostly because so many companies - despite the GDPR - have rudimentary, manually when not completely nonexistent processes to delete your data.

In this post I describe my process going through about 100 old accounts and trying to delete them all, including a top 10 for the weirdest, funniest or most interesting cases I encountered while doing so.

Wilder7977··on How we rooted Copilot
Lichess, the gazillion of self-hosting software. There are many examples of free software that are exclusively (or let's say predominantly) used in noncommercial environments.

In any case, I agree with the commenter, and I think that developing a software which is also used by companies is different from looking for vulnerabilities in the context and scope of a bug bounty program for a specific company. Yes, you could argue that users of said company are going to be more secure, but it's evidence t like even in this case the company is the direct beneficiary.

Wilder7977··on Do not download the app, use the website
A good strategy for this is to enroll it at day 0 for the most sensitive systems (e.g., password manager, email accounts). This way you are able to use it as a backup in the sense of giving the option to reset or access (e.g., via backup codes) all the services, without being necessarily enrolled in all of them.
Wilder7977··on Do not download the app, use the website
I wish that was an option, in most cases the phone becomes the hardware token, and that can be lost too. Or broken, or out of power or without internet connection.

I even have a personal anecdote. My wife "lost" her phone in Iceland. I make her login to find-my-phone with her google account, and 2fa was needed. Thankfully she had her Yubikey in her keychain (plus, we enrolled each other's key), so she was able to login. Push notification or TOTP/SMS were all not an option.

Wilder7977··on Reflections on OpenAI
Allow me to propose a different rationalization: "yes I know X might damage some people/society, but it was not me who decided, and I get lots of money to do it, which someone else would do if not me."

I don't think people who work on products that spy on people, create addiction or worse are as naïve as you portrayed them.

Wilder7977··on Million Times Million
Small typo, but it's "Miliardo" (one "l").
← PreviousPage 3 of 4Next →