From the article it seems the name is 马启仁, not 马骑人 so the guy's name sounds the same as 'horse riding man', but that's not a literal translation of his name.
I wonder how good a job the ClearScan feature on Adobe Acrobat would do. IIRC it creates one or more fonts based the existing characters in the PDF. So each lowercase 'a' would look the same, and be a sort of average of all the 'a' letters in the book.
Interesting – I have some old British academic press books with "advanced" typographical features; I'll see how it does with them. (The type you have to break out advanced typography features in Adobe apps to set – old style and modern numerals, interpuncts, two types of ampersands - one for headings and body text, swash alternate forms in titles.)
If it could do anything that a junior dev could, that’d be a valid point of comparison. But it continually, wildly performs slower and falls short every time I’ve tried.
Trying to make a media player, media server, all by using ffmpeg and a pre-built media streaming engine as it's core. Python and SQLite. About a week's worth of effort every time until it begins to go too far off the rails to be reliable to continue to develop with. It never did get the ffmpeg commands right, I had to go back to crafting those by hand, it never did get the streaming engine to play in the browser's video player in the supported hls and dash formats. Asked it to build a file and file metadata caching layer and then had to continue to re-prompt it to poll the caching layers before trying to get values from the database. Never even got to the library, metadata, or library image functionality. Had to ask it to create the rbac permissions model I wanted despite it being very junior-level common sense (super-admin, user-admin, metadata admin, image admin).
I recently built something in the same universe - using ffmpeg to receive streams from obs to capture audio and video - don't want to get into details beyond except to say it involved a fairly involved pipeline of ray actors and a significant admin interface with nicegui. I had no problem doing this with claude. You need to give it access to look up how do things, like context7. If you are doing something very specific, you need to have a session that does research to build a skill so it doesn't need to redo that research every time. And yes, you do need to tell it the architecture and be fairly detailed with something like how you want rbac.
Using these tools takes quite a bit of effort but even after doing all those steps to use the tool well, I still got this project done in a few days when it otherwise would have taken me 1-2 months and likely simply would never happened at all.
I'm curious which harness and which model(s) you've been using.
And whether you have a decent PRD or spec. Are you trying to prompt the harness with one bit at a time, or did you give it a complete spec and ask it to analyze it and break it down into individual issues with dependencies (e.g. using beads and beads_viewer)?
I'm not looking for reasons to criticize your approach or question your experience, but your answers may point to opportunities for you to get more out of these tools.
> A. You're working on some really deep thing that only world-class expects can do, like optimizing graphics engines for AAA games.
This is a relatively common skill. One thing I always notice about the video game industry is it's much more globally distributed than the rest of the software industry.
Being bad at writing software is Japan's whole thing but they still make optimized video games.
It’s a simple compiler optimization over bayesian statistics. It’s masters-level stuff at best, given that I’m on it instead of some expert. The codebase is mixed python and rust, neither of which are uncommon.
The issues I ran into are primarily “tail-chasing” ones - it gets into some attractor that doesn’t suit the test case and fails to find its way out. I re-benchmark every few months, but so far none of the frontier models have been able to make changes that have solved the issue without bloating the codebase and failing the perf tests.
It’s fine for some boilerplate dedup or spinning up some web api or whatever, but it’s still not suitable for serious work.
When really solid programmers who started skeptical (and even have a ban policy if PR submitters don’t disclose they used AI) now show how their workflows have been improved by AI agents, it may be worth trying to understand what they are doing and you are not.
Claude would be worse than an expert at this, but this is a benchmarkable task. Claude can do experiments a lot quicker than a human can. The hard part would be ensure that the results aren't just gaming your benchmark.
Companies are not comparing it straight to juniors. They're more making a comparison between a Senior with the assistance of one more more juniors, vs a Senior with the assistance of AI Agents.
I feel like comparison just to a junior developer is also becoming a fairly outdated comparison. Yes, it is worse in some ways, but also VASTLY superior in others.
It’s funny so many companies making people RTO and spending all this money on offices to get “hallway” moments of innovation, while emptying those offices of the people most likely to have a new perspective.
"I think political interference is a horrible thing for university education."
The University of California is one of the largest universities in the US. It is governed by a Board of Regents. The majority of those Regents are appointed by the state Governor.
Do you consider that 'political interference'?
One of the things those Regents did was vote to end the use of SAT scores in admissions. They did during a meeting in which several spoke of the value of the SAT. And they acted against the recommendations of the Academic Council's Standardized Testing Task Force.
You might think that the staggered and long terms protect against political interference/influence. But if that's the case, how do we explain how so many votes are unanimous when, on the day of the vote, some regents express opposing views?
Boards of Regents consistently suck shit. The rather famous "put your body upon the gears" speech was about the Berkeley Board. Leftists largely hate the boards of both public and private universities. They are often megarich people with minimal understanding of pedagogy or even university administration.
> how do we explain how so many votes are unanimous when, on the day of the vote, some regents express opposing views?
That reminds me of the Politburo voting scene in The Death of Stalin. Small group politics at their finest.
Anyway, the UC Board of Regents is full of political hacks and corrupt cronies. Diane Feinstein's husband was famously a regent, while simultaneously serving as Chairman of both CBRE and his own leveraged buyout private equity firm.
It's also illegal in many jurisdictions (e.g. in the US, viewed as a scheme to defraud advertisers by generating invalid clicks that cause financial harm, by depleting their budgets and push them to spend for fake traffic), but in practice it's way easier to just blacklist that IP / user.
The big networks filter such traffic, the small networks benefit from it.
You may also get accidentally get your own website blacklisted or moved to a lower RPM tier, or provoke shadow-ban websites that you like to visit, or... generate more ad revenue for them.
Don't tell me I'm not allowed to click buttons you put in my face.
Any jurisdiction where this is supposedly illegal, it hasn't been court tested seriously.*
Per your link: "What you're describing is essentially the extension AdNauseam. So far they have not had any legal troubles, but they technically could." That stance or an assertion it's not illegal is consistent throughout the thread, provided you aren't clicking your own ads.
"The industry" thinks you shouldn't be allowed to fast forward your own VCR through an ad either. They can take a flying .. lesson.
* Disclaimer: I don't know if that's true, but it sounds true.
Some years ago I was by chance listening to a radio program about advertising. They interviewed a marketing guy and he insisted that it was illegal for you to visit the bathroom or the kitchen while the ad was running (on TV or on the radio). Completely nuts.
That reminds me of the time I was flipping through TV channels and stopped in on TBN to see what color Jan's hair was going to be. Instead, I found Paul preaching about how anyone watching his programming and NOT sending him donations was stealing from him.
>Don't tell me I'm not allowed to click buttons you put in my face.
No, the illegal-ness doesn't come from the clicking, it comes from the fact you're clicking with the intention of defrauding someone. That's also why filling out a credit card application isn't illegal, but filling out the same credit card application with phony details is.
The intent isn’t to defraud. The intent is to curb their uninvited data collection and anti-utility influence on the internet.
You’re not defrauding anyone if you have your extension click all ads in the background and make a personalized list for you that you can choose to review.
>The intent isn’t to defraud. The intent is to curb their uninvited data collection and anti-utility influence on the internet.
How's this any different than going around and filling out fake credit applications to stop "uninvited data collection" by banks/credit bureaus or whatever?
>The intent is convenience and privacy, not fraud.
You're still harming the business, so my guess would be something like tortious interference.
In a credit application there is a signature and binding contract. If I fill in false information knowingly, the intent is clear and written.
If you send me an unsolicited mailer with a microchip that tracks my eyes and face as I read it, you’ve already pushed too far. To then claim my using a robot to read it for me is fraud ignores the invasion of privacy you’ve already instituted without my express consent (digital ads are this).
It’s not fraud if it’s self-defense from corporate overreach.
I am super curious how far this goes. If, hypothetically, I wore some sort of glasses that kept facial recognition from identifying and tracking me at my local grocery store, would that constitute a civil infringement in the future?
What about extensions that skip embedded ads in a YouTube video? Is that tortuous interference with the view counter that creators use to market their reach?
>How's this any different than going around and filling out fake credit applications to stop "uninvited data collection" by banks/credit bureaus or whatever?
It's so different that it can't even be compared. There's nothing similar there.
>>The intent is convenience and privacy, not fraud.
> You're still harming the business, so my guess would be something like tortious interference.
No, you're not harming the business. You're simply not following the business idea of the "business". Anyone can have a business idea of some type. Not a single person on earth has any obligation to fulfill that business idea. But somehow some people believe the opposite.
> No, the illegal-ness doesn't come from the clicking, it comes from the fact you're clicking with the intention of defrauding someone. That's also why filling out a credit card application isn't illegal, but filling out the same credit card application with phony details is.
You might technically be right. But I'd recommend contacting EFF, if, somehow, installing AdNauseam brings you into legal trouble.
On the realm of search engines and ad networks I love to remind people that Google took out "don't be evil" from their motto and pressured anyone within US jurisdiction to remove Page and Brin's appendix #8 (at the least it's removed from their original school of Stanford).
1: Ad companies are not going to go after individual users, rather they would target the maker of any such plugin
2: If they did go after an individual user, they would have to prove damages, and an individual is unlikely to do more than a few bucks of wasted ad spend for a company, not even a rounding error, making the legal cost and political cost of targeting the person running the script enormous compared to the potential return from anything other than a grand slam nuclear judgement in their favor.
1) The makers of this plugin are from EFF, and thus have the time and resources to combat litigation.
2) Yep! And as mentioned in other threads, it would give the users on their ad platform more money but degrade the quality of their ad platform.
I was just alarmed by how many people are not only okay with, but defending, the current state of ad tech. I think it's a noble effort to go against the grain and withstand any potential legal trouble to subvert it as it seems there's no recourse to be made in the courts unless an entity has the aforementioned time and money to fight it in the courts.
stanford.edu, and the appendix is there. In fact on the link you gave the appendix is cut short - looks like an OCR/copying issue but then at a glance it doesn't seem to happen elsewhere which is a little suspicious. I'm not sure what you're talking about.
I must have somehow missed that one; glad that ancient site without HTTPS is still up. Here are the two top results I get from searching for it from Stanford[0][1], and you can see that this section of the appendix is missing. Google's also has it missing[2]. So no, I don't think I'm crazy.
Touché! I recant my conspiratorial thinking. Though I still think it's odd that the other sources I posted don't have it; one is what's actually being taught in Stanford courses and the other is Google's own hosting of their founders' paper.
You'd be doing way more harm than good. The battle between ad networks and unscrupulous website owners using bots to fake ad clicks has been going on forever.
I don't think the question was about whether this would actually help the advertisers. (I suspect it was rhetorical.) Of course the defense will now be harder to execute for anyone who reads this thread.
>People purchase visibility and clicks when they purchase advertising. not conversions or sales.
Again, you're ignoring intent in all of this. It's not illegal to default on a loan, or even to refuse to pay it back (eg. bankruptcy), but it is illegal to take out a loan with the specific intent to not pay it back (eg. if you know you're planning on declare bankruptcy right afterwards).
I mean, (not to you, as we go in the same direction, in general), just block it.
The goal of Adnauseam was to hurt Google, and other big adnetworks, from what I understand.
By blocking:
-> Advertiser is not harmed
-> For the adnetwork: No ad revenue
-> Publisher is not harmed
-> Pages load faster
--> Google is earning less (if this is part of your ideological fight) and you get rewarded with a better experience, and you are legally safe
==
With fake clicks:
-> Advertiser is harmed
-> Publisher is harmed
-> Adnetwork is okayish with the situation (to a certain point)
-> You hurt websites and products that you like (or would statistically like)
--> Google is accidentally earning more revenue (at least temporarily, until you get shadow-banned), your computer / page loads slows down and you enter a legally gray area.
(+ the side-note below: clicking on every ads leak your browsing history because in the URL there is a unique tracking ID that connects to the page you are viewing)
I still wonder about that. I don't have a contract with the advertiser to provide genuine data back about what ads I've clicked and what I haven't. The website operator does have such a contract and so cannot hire a bot farm to spam click the ads.
If it's something that's been held up in court already then of course I have to accept it, but I can't say the reason seems immediately intuitive.
>I don't have a contract with the advertiser to provide genuine data back about what ads I've clicked and what I haven't.
Charges of fraud doesn't require a contract to be in place. That's the whole point of criminal law, it's so that you don't need to add a "don't screw me over" clause to every interaction you make.
Wrong. There is no law saying you cannot click every link on a website within your browser. It would not only be impossible to prove but also entirely wrong interpretation of existing laws.
Now if you had an AdWords account and ran a botnet that visited your property and clicked ads, that’s fraud.
Back up a bit. AdNauseam and similar tools are not illegal. The only real avenues would be violation of ToS, fraud, computer abuse or similar. For an individual running this on their home PC for their own use it would be a real challenge for anyone of any size to prove harm.
Now like I already said, if you are running a botnet clicking on your ads that is entirely a different story.
So tell us what does having the extension installed prove?
click fraud consists of the person who runs a website themselves clicking, running bots to click, paying someone else to click, etc ads on their own website. it becomes fraud first because they have contractually agreed not to do that, and second because they are materially benefiting from it. an unaligned third party clicking (etc) on ads has neither of those conditions being true, and hence isn't fraud or otherwise illegal.
If you intentionally loop-download large files or fake requests on websites that you don't like, in order to create big CDN charges for them, then what ?
Without reaching the threshold of Denial of Service, just sneakily growing it.
Nobody benefits, except for the weird idea of the pleasure of harming people, still illegal.
> The opinion states: “click fraud” can occur when “either a (natural) person, automated script, or computer program, sometimes referred to as a `bot,’ simulates the click activity of a legitimate user by clicking on the Program Data displayed, but without having an actual interest in its subject matter or content.”
You're all over this thread spreading misinformation. AdNauseam has been around since 2014. It is specifically banned in the Chrome store so Google knows of it's existence. If you check the wikipedia page you'll see that they have landed in the press and taken multiple actions against the extension.
https://en.wikipedia.org/wiki/AdNauseam
Usually when it's brought up people say it doesn't work or try to spread fear that it is illegal. Google banning them but taking no action otherwise indicates to me and the thousands who use it that it is in fact effective and Google has no other recourse other than their control over the most popular browser.
You deliberate harm and financial damage using a computer bot. Almost all countries have provisions where you can be sued for any type of damage you cause and be asked to repair it (a minima at the civil level).
Big ones detect it, so they don't care to sue. Small ones benefit, so they don't sue.
This is your main protection, there is nothing to squeeze from a single guy. Even if you get him to pay you back the fraud, then what ? It costs more in legal fees.
Still, it's such an odd concept to self-inflict yourself such; it's way better to just block the ads than to be tagged as a bot and get Recaptcha-ed or Turnstiled more frequently.
> It's a very harmful practice to intentionally try to hurt companies, when you can just block what you don't like.
I say tit for tat. They're intentionally trying to harm me, spying on me, maybe infecting my computer, mining crypto with my CPU, or wasting my network bandwidth. They could just not do that and there wouldn't be any concern about reciprocity
Does your say have any relevance here in terms of what the law is? Are you a state judge tasked with interpreting the law? Where's the tit-for-tat clause?
This is not ok I totally agree with you, but still, I would rather just block the ads, and not buy their products or support them.
There is a side-effect in terms of privacy: you send a fake click request every single time, you also actually disclose to adnetworks which page you are visiting and incidentally your whole browsing history (not through referrers, but because click URLs have a unique click IDs to match).
I've never understood the use-case of Adnauseam. This just, essentially, allows the adbroker (e.g. Google) to get more money from the business putting up the ad. Unless every single person uses it, it's not going to stop business from advertising, it just makes the likes of Google get more revenue.
>> This just, essentially, allows the adbroker (e.g. Google) to get more money from the business putting up the ad.
It lowers the effectiveness of internet advertising. When advertisers feel they're paying too much for the business the ads generate, they'll stop advertising in that way. That's probably the thinking anyway. A less generous stance would be: I hate advertisers so I'm gonna get back at them by making them pay more.
It would just cut the rates they'll pay to account for the erroneous clicks. I guess that might just be limited to defunding the sites popular with the really techy group of people that use Adnauseam and instead shift to niches with better effectiveness.
Assuming it actually works (which I'm not sure about), it increases the cost on the business putting up the ad (presumably targeting you). It acts as a small punishment to the business buying the ads I guess.
>Assuming it actually works (which I'm not sure about),
Which it probably doesn't, given that it uses XHRs to "click" on ads, which is super detectable, and given the proliferation of ad fraud I'd assume all networks already filter out.
The other assumption here is that ad networks want to filter out all clicks but the most legitimate.
I don't think that's a very lucid assessment of how advertisers operate on the Internet. We all agree that they could take these steps. If AdNauseam doesn't look like outright fraud in the logs (which they don't if it's all distinct IPs and browsers), I don't think they want to cut it out from their revenue and viewer analytics.
it's actually the opposite, google adsense and every major ad-network will ban you or put a hold on your account if they think the ad impressions or clicks are automated, so this is a good way to get someone blocked from the ad-network
Makes your laces sit a bit funny compared to regular
The 'sit a bit funny' issue is the classic symptom of 'the granny knot'.
If you have inadvertently been tying granny knots, you may notice:
1) Instead of the bows hanging to the sides, they naturally want to hang along the length of your show (one pointing diagonally away from you, and the other diagonally towards you).
2) Your shoelaces get undone often, unless you do a double knot.
The fix (whether you tie your bow using the regular way, bunny ears, or Ian Knot) is to reverse the direction of your initial knot.
If you watch this video I made, you will see that the Ian Knot (when done according to the instructions on Ian's site) results in the laces sit just how they should: https://youtu.be/JaBmehtalAY
> The fix (whether you tie your bow using the regular way, bunny ears, or Ian Knot) is to reverse the direction of your initial knot.
Far too many people say that you need to reverse the direction of your main knot. This also works, of course, but it's way more difficult to unlearn then relearn the main knot. Far easier to change the direction of the initial knot. When I first learn the Ian Knot I quickly discovered I'd learnt it "backwards". So I reversed the initial knot and I've been tying it that way for close to 20 years now.
I found only two videos about it on YouTube. This is the better of the two, and illustrates the output: https://www.youtube.com/watch?v=vndn2vmSIbw
I don't know whether that video uses Kumo (UI library also from Vercel).
reply