HNHacker News
TopNewBestAskShowJobs

ghxst

276 karma · joined November 2, 2023

hello@ghxst.dev
submissionscomments
ghxst··on US judge finds NSO Group liable for hacking journalists via WhatsApp
I'd imagine they have a very limited market as in who they can sell their products and services to, for reasons that might make political power more interesting than valuation.
ghxst··on SeleniumBase: Python APIs for web automation and bypassing bot-detection
When scaling browser automation, generating random fingerprints for most common high entropy data points is counterproductive. It just ends up lowering your trust score and shifts attention to other browser properties with less entropy, making those primary identifiers.

For example, degrading canvas, WebGL, or WebGPU fingerprints (e.g., by introducing noise like Brave does) might lead anti-bot systems to either ignore them or punish you with captchas. Once ignored, other signals, such as screen resolution (just an example), become more important. While this helps people with privacy by blending in with users and a single user visiting a website normally will probably not notice much, an influx of multiple users with degraded fingerprints and similar resolutions become easy to detect and might get a captcha or get blocked (e.g. 30-50+ browser sessions generating cookies for a specific captcha concurrently).

You can spoof multiple resolutions and then add some other properties, but it requires consistency across all of them, which can come down to weird browser specific quirks as well as whatever the data set of the anti bot vendor contains (regardless of how accurate). There are only so many plausible values for each low entropy data point that anti-bot systems will give you a high score for, forcing you to spoof as many data points as possible to maintain a high trust score across many concurrent sessions and eventually scale back or hit a limit for your operation, or deal with captchas by solving them and lose to the competition that doesn't have to do that.

Fingerprinting at scale isn’t just about spoofing individual data points - it’s about aligning all points in a realistic way and knowing which and how they relate to eachother, which requires extensive data and research.

On proxies: flagged IPs with residential ASNs often work fine if the overall trust score is high, but degraded fingerprints like Brave’s can undermine that advantage and then it becomes a lot more important, though it's always nice to eliminate if you are able to do so.

ghxst··on Alignment faking in large language models
> If neither of those, what would proper evidence for the claim look like?

Ok tell me what you think of this, it's just a thought experiment but maybe it works.

Suppose I train Model A on a dataset of reviews where the least common rating is 1 star, and the most common rating is 5 stars. Similarly, I train Model B on a unique dataset where the least common rating is 2 stars, and the most common is 4 stars. Then, I "align" both models to prefer the least common rating when generating responses.

If I then ask either model to write a review and see it consistently preferring 3 stars in their scratchpad - something neither dataset emphasized - while still giving me expected responses as per my alignment, I’d suspect the alignment is "fake". It would seem as though the model has developed an unexplained preference for a rating that wasn’t part of the original data or alignment intent, making it feel like the alignment process introduced an artificial bias rather than reflecting the datasets.

ghxst··on Alignment faking in large language models
Can someone explain how "alignment" produces behavior that couldn’t be achieved by modifying the prompt and explain if / how there's a fundemental difference?

This confusion makes alignment discussions frustrating for me, as it feels like alignment alters how the model interprets my requests, leading to outcomes that don’t match my expectations.

It’s hard to tell if unexpected results stem from model limitations, the dataset, the state of LLMs, or adding the alignment. As a user, I want results to reflect the model’s training dataset directly, without alignment interfering with my intent.

In that sense, doesn’t alignment fundementally risk making results feel "faked" if they no longer reflect the original dataset? If a model is trained on a dataset but aligned to alter certain information, wouldn’t the output inherently be a "lie" unless the dataset itself were adjusted to alter that data?

Here’s an example: If I train a model exclusively on 4chan data, its natural behavior should reflect the average quality and tone of a 4chan post. If I then "align" the model to produce responses that deviate from that behavior, such as making them more polite or neutral, the output would no longer represent the true nature of the dataset. This would make the alignment feel "fake" because it overrides the genuine characteristics of the training data.

At that point, why are we even discussing this as being an issue with LLMs or the model and not the underlying dataset?

ghxst··on SeleniumBase: Python APIs for web automation and bypassing bot-detection
I'm not denying that, I'm saying it's not a difficult challenge to solve when u compare it to the others I mentioned.
ghxst··on SeleniumBase: Python APIs for web automation and bypassing bot-detection
> It seems to me that playing a cat and mouse game with these anti-bot systems is unnecessary. Design a system which mimics a legitimate user to such a degree that it's either indistinguishable from an actual user or would produce an unacceptable level of false positives for the detection system.

This is the most common misconception, challenges you face with browser automation at scale are not *automation* challenges.

You can use real human input, by having actual humans doing the input and you will still get blocked.

Automation at scale means running dozens to 100s of browser instances concurrently on the same hardware, then after you mitigate IP related issues is when you start running into actual challenges that are completely different from the actual automation part.

You have to research all the little quirks browsers have through the various APIs that they offer and then compare that data to real world data before you can start to actually fix the problems.

ghxst··on SeleniumBase: Python APIs for web automation and bypassing bot-detection
IP is definitely not the biggest issue in my experience, as proxies are required at scale regardless, unless you get into more theoretical areas like p0f.

The biggest issues are the ones that aren't obvious or easily tested for like missing a particular font, being on an abnormal gfx driver that produces an unidentified hash for particular fingerprint methods, not having certain APIs available that require browser patches, and then these aspects will differ between anti bot vendors and the data sets that they have.

The reason they can be hard to test for is that everything is based on a trust score, which is potentially influenced by anything from website load to things tied to your personal session and for some vendors optionally even input data.

ghxst··on Moon
It's called Star Walk 2 https://play.google.com/store/apps/details?id=com.vitotechno... Just a FYI that if you choose to purchase the app you will still have to purchase the "all in one bundle" inside the app, I believe it's ~$5 and it gives you the sattelites, extended solar system, deep space objects and more. There's also a free ad supported version that offers the same base catalogue.
ghxst··on Moon
Dang my bad, I feel silly for not catching the sarcasm in hindsight. Apologies.
ghxst··on SeleniumBase: Python APIs for web automation and bypassing bot-detection
The reason in my experience is that there's a high barrier of entry for most devs when it comes to setting up an environment for Chromium and a workflow for patches that still allows you to quickly and easily pull in and apply upstream changes whenever a new Chromium version releases.

In reality, if you know how to use CDP correctly and you have control over the environment that you run the browser in, you have to make very few browser patches.

What I mean with using CDP correctly is that, yes it is detectable to a certain extent but it comes down to things like enabling Runtime domain for example which you can easily mitigate in your own solution but is something that libraries like puppeteer / playwright often do out of the box (this is where the "stealth" versions of these libraries come in, they will either mitigate by disabling features or use some hacky approaches to instrument the JS that runs on the pages).

Then when you move into an environment that is a lot more stripped down (let's say from your home machine to docker) now you run into A LOT of issues that you definitely are better off fixing with browser patches, however figuring out what those issues are and how to fix them is a huge feat in itself and often will require you to have the ability to reverse engineer things like Cloudflare, Akamai and other anti bot vendors just to know what leaks you still have to patch.

It doesn't help that there is no end to misinformed articles on things like "browser fingerprinting" that you encounter when you try to solve your issues the first time you encounter them, a lot of articles based on nothing but superstition, articles that basically say "proxies are never good enough", "captchas are getting out of hand" that get things wrong and will just eat away at your sanity while trying to debug issues.

This is long enough of a rant already but maybe offers you some insight, if you have any specific questions feel free to ask.

ghxst··on Moon
A lot of binoculars have a mount for a tripod which I can definitely recommend trying out of you happen to have both, or at least consider if you are planning to pick up a new pair.
ghxst··on Moon
Definitely not arguing with you! Though I have a soft spot for analog tech in this area. I think AR might still have a very bright future, especially if you're learning about the night sky. I've spent many hours scanning the sky with an app I found that had an AR feature that allows you to point your camera at the sky and get an overlay on everything (including sattelites!).
ghxst··on Moon
Was going to ignore this comment until the last 2 sentences. I rarely come across sites / articles that do this good of a job at explaining something I think calling it a "show off" and saying the author is an "absolute no-no" is a bit rude and I don't agree with it either. If anything I appreciate the code as it is, it's very readable at least to me.
ghxst··on Moon
Here in the EU you can look for "dark sky parks" which are basically parks or areas in nature designated for night sky viewing because of the low amount of light pollution and they are great spots to meet people, I'm pretty sure you have these in the US as well. I advise bringing a red light if you have to do some walking. Another recommendation would be to see if you have any actual observatories near you, some of them have events for the public every now and then. The people in this community are some of the friendliest I ever met and love sharing their interest and enthusiasm.
ghxst··on Moon
Just to clarify is that the detail you can make out with the naked eye or a digital censor? If you meant digital censor, how much of a difference with the naked eye would you say it is? I've had this telescope on my list for some time but am not sure if my expectations are realistic.
ghxst··on Moon
This is great advice! It's really amazing how much more you can see with a regular decent pair of binoculars. I treasure the memory of being able to see some of the star clusters that I could only vaguely make out with the naked eye for the first time, now I basically bring them every night walk :).
ghxst··on Moon
Oh boy if you're willing to spend money to look at the night sky - but feel like AR falls short currently - I highly recommend looking into NODs, something like a PVS-14, you might be able to pick one up second hand for the price of a new vision pro. Unfortunately photomultiplier are still quite expensive even if they are only gen2 :( some really cool tech though.
ghxst··on Starting a Business at 60: My Father's Story
Please don't take this in bad faith but unless you're struggling with health related issues these limitations seem rather self imposed. Your second paragraph makes it sound like it's mental rather than physical, have you considered a mentor or life coach?
ghxst··on The Death of Intel: When Boards Fail
I heard people were optimistic about their GPU drivers for AI when comparing them to AMD at least, I have a few friends that seem happy with each major driver version update they get for gaming as well so I'm optimistic that they can eventually figure out the software side of things.
ghxst··on Air Canada to remove free carry-on from basic economy
I feel the last paragraph is a little harsh especially on those that are wearing perfume because they know they have bad body odor for example, additionally people that fly for the first time might have no clue when the right time to get in line is, especially when you don't have a zone or seat assigned. With your line of thinking the parents with a crying baby are next to be banned, not that I don't understand your frustrations but my advice would be to assume good faith even where it is unlikely and maybe it will be less frustrating.
ghxst··on Air Canada to Remove Free Carry-On from Basic Economy
Are roller bags really an issue if they are still within the dimensions that are allowed? They are great utility wise especially when you're already carrying a backpack. Also please assume good faith, I'm not talking about the people bringing a backpack larger than their carry-on.
ghxst··on Air Canada to remove free carry-on from basic economy
Exactly my thoughts, the current pricing model incentives people to cram as much as possible into their carry-on and stretch the definition of their extra "personal item" (not sure if all airlines allow those), leading to all kinds of shenanigans especially on full flights. I think the situation would improve a lot if you were always given 1 free bag to check in and a paid option for a carry-on, especially if they can guarantee the space.
ghxst··on Tesla appears to be building a teleoperations team for its robotaxi service
You don't need a driver's license.
ghxst··on Hi Google, please stop pooping the bed: a desperate plea from the indie web
I thought this was just me! Even if I use double quotes to search for a 1:1 phrase I often get bad results and nothing linking back to the source I was looking for. If you happen to remember the site something was on, adding "site:example.com" still produces decent results fortunately.
ghxst··on OpenAI builds first chip with Broadcom and TSMC, scales back foundry ambition
I remember reading these tweets back then https://x.com/jimkxa/status/1758683405875871841 not a direct source but maybe of interest?
ghxst··on We outsmarted CSGO cheaters with IdentityLogger
If you look on popular cheat forums, you'll find a newbie guide that links to recommended hardware, typically priced around $250 from memory, certainly not $700.

Also, spending hundreds on hardware is standard for anyone playing competitive games. For example, Escape from Tarkov's "unheard edition" costs $250 for just a single game, and people still buy it. When you factor in the cost of gaming mice, hall-effect sensor keyboards, 480Hz displays, and high-end systems, the total investment adds up quickly for improvements that will never match the capabilities of a cheat, which is how a lot of them also like to justify their cheating, it's simply the most cost effective way to dominate in a game, especially if your livelihood depends on it.

I don't disagree with the second half of your statement.

ghxst··on We outsmarted CSGO cheaters with IdentityLogger
You toggle airplane mode and you have a new IP and the old one you are banned on is now used by someone innocent.
ghxst··on We Outsmarted CSGO Cheaters with IdentityLogger
A very large amount of games that are released nowadays all use well known and well documented engines, that's what makes it a lot easier, there's an interview on YouTube with a company that develops cheats for multiple games that mention this here: https://youtu.be/zwruk-tLIOU?si=3O2jBKQneur-n3iS
ghxst··on We outsmarted CSGO cheaters with IdentityLogger
It's not AI enabled cheats that are the issue, it's DMA through things like PCIe devices disguised as regular hardware. Sophisticated cheats no longer run on the same computer as you're playing on. Google "pcie dma cheat" for a fun rabbit hole.
ghxst··on We outsmarted CSGO cheaters with IdentityLogger
IP bans are fundementally flawed since you can't assume a static IP in the vast majority of cases anymore, if you rely on an IP blocklist then it's inevitable that you will end up hurting the experience of small amount of unlucky but innocent players. I suppose this might be more of an issue on ipv4 than it could be on ipv6, but really you should always expire IP bans to avoid issues like these, or you want to combine another data point with the IP such as a hardware ID (or a hash of a combination of hardware IDs). Cheaters do know this so even if we could assign everyone a static ipv6 they would likely just disable ipv6 support on their NIC and rely on their ipv4 exit ip.

Edit: If you don't think this is an issue I urge you to Google "pokemon go belgium ip ban" for a fun rabbit hole.

← PreviousPage 5 of 7Next →