frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Gemini 3.8 Flash and 3.8 Flash Cyber

https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-c...
511•bratao•3h ago•299 comments

A Note from LWN

https://lwn.net/Articles/1090585/
592•rwky•5h ago•122 comments

I Don't Have a Smartphone

https://ploum.net/2026-09-02-i_dont_have_a_smartphone.html
46•speckx•40m ago•35 comments

Paint.net 5.2 alpha now runs on Linux

https://forums.paint.net/topic/134562-paintnet-52-alpha-build-9739/
53•judah•1h ago•25 comments

Mistral now trains on user input by default, except on enterprise tier

https://help.mistral.ai/en/articles/455207-can-i-opt-out-of-my-input-or-output-data-being-used-fo...
275•teekert•6h ago•122 comments

Biggest dark matter detector spots a single weird particle

https://www.science.org/content/article/world-s-biggest-dark-matter-detector-spots-single-weird-p...
168•randycupertino•4h ago•39 comments

GrapheneOS says Pixel 11 has MTE support after all

https://grapheneos.social/@GrapheneOS/117194007157499435
145•user_7832•4h ago•91 comments

Exit the Cave

https://turtlespace.blog/p/exit-the-cave
126•akkartik•4h ago•26 comments

A third of Perplexity's citations don't contain the number they're cited for

https://hausresearch.com/reports/perplexity-citation-audit/
105•jakobgreenfeld•4h ago•37 comments

Three sites made 215,128 “best software” pages for AI. Perplexity cites them

https://trellner.com/reports/manufactured-sources-behind-ai-recommendations/
206•jakobgreenfeld•4h ago•101 comments

SteamdDB Joins Nexus Mods

https://www.nexusmods.com/news/15597
46•HelloUsername•3h ago•16 comments

A Selection of Los Alamos Rolodex Business Cards

https://clui.org/collections/los-alamos-business-cards/selection-cards
90•1970-01-01•1d ago•19 comments

Aging Brains Blend Memories Together Instead of Just Forgetting Them

https://studyfinds.com/aging-brains-blend-memories-together-instead-of-forgetting-them-study-finds/
117•mdp2021•5h ago•48 comments

Commodore 64 released September 1, 1982

https://dfarq.homeip.net/commodore-64-released-september-1-1982/
276•giuliomagnifico•9h ago•134 comments

Poisson Disk Sampling

https://stripeacross.com/posts/poisson-disk-sampling/
86•vismit2000•4h ago•12 comments

Making the Internet Boring

https://cemrehancavdar.com/2026/08/30/making-the-internet-boring/
15•zdw•2d ago•2 comments

WebLLM: high-performance in-browser LLM inference engine

https://github.com/mlc-ai/web-llm
56•saikatsg•4h ago•11 comments

Show HN: FrontierHarness Eval – 9 harness, same model, cost per pass varies 17x

https://frontierharness.org
40•shiqimei•2h ago•16 comments

Why do so many tools have JSON config files?

https://textlog.cc/post/895
3•stagas•9m ago•0 comments

Six curl CVEs after OpenAI and Anthropic came back with zero

https://aisle.com/blog/aisle-discovered-six-curl-cves-after-openai-and-anthropic-found-zero
129•goobreee•4h ago•44 comments

Discontinuation of third level domain registrations for the .name TLD [pdf]

https://itp.cdn.icann.org/en/files/consensus-policies/rsep-2026013-name-request-15-04-2026-en.pdf
26•greyface-•1d ago•11 comments

How to debloat your Xiaomi 15 Ultra without root access

https://trackerninja.codeberg.page/post/how-to-debloat-your-xiaomi-15-ultra-without-rooting-and-c...
16•spacedrone808•1h ago•9 comments

Tangle – Visual ML Pipeline Editor

https://tangleml.com/
11•duck•1h ago•1 comments

Quasar 438B: Europe's Leading AI Model

https://multiversecomputing.com/resources/introducing-quasar-438b-europe-s-leading-ai-model
143•amunozo•8h ago•94 comments

Check if a file was made with Claude

https://claude.com/check-content
108•frexs•6h ago•77 comments

Telli (YC F24) is hiring engineers and designers [Berlin, on-site]

https://careers.telli.com/
1•sebselassie•9h ago

LLMs: Intelligence vs. Cost

https://openteams.com/intelligence-vs-cost/
58•theanonymousone•5h ago•24 comments

The Emergent Symbolic Structure of Artificial Neural Networks

https://arxiv.org/abs/2608.29530
260•schmuhblaster•14h ago•88 comments

How Railroad Crossings Work

https://practical.engineering/blog/2023/12/29/how-railroad-crossings-work
15•at1as•1h ago•6 comments

The race to engineer new knobs for the human brain

https://www.neuroai.science/p/the-race-to-engineer-new-knobs-for
13•pminimax•2h ago•0 comments
Open in hackernews

ChatGPT ad targeting is garbage

https://successfulsoftware.net/2026/09/02/chatgpt-ad-targeting-is-garbage/
33•hermitcrab•1h ago

Comments

Insanity•1h ago
Ads are part of the enshitification of ChatGPT. In the short-term it might offer them some kind of lifeline and I'm sure in their next investor call they'll highlight their revenue from Ads.

But if their competitors (including open-weight models) can offer similar quality and ease of use, Ads is just another reason _not_ to use them.

lnsru•58m ago
I saw them today in the app and yeah… first idea - enshitification. Ads were cringe. But google works the same way - cringe ads. I use google in 3-4 different languages and it’s completely lost.
ryandvm•32m ago
Ads are responsible for the enshitification of everything. Literally everything.

- Why is online news just partisan rage-baiting? Ads

- Why is real journalism dying? Ads

- Why are your web search results less helpful? Ads

- Why are recipes 2,000 words long? Ads

- Why is social media engineered for scroll-maxxing?

- Why is your car/TV/refrigerator spamming you?

- Why are pharmaceuticals so expensive?

- Why is your population voting for assholes?

Future generations are going to look back at how we live with advertising the same way we look back on leaded gasoline and asbestos.

livinglist•26m ago
I occasionally enjoy well executed ads, like something that’s creative enough to make me curious, but nowadays, I’m seeing a lot of ads on YouTube that’s obviously low-effort AI slop, they are boring, fake, and depressing.
beezlebroxxxxxx•14m ago
> Why is real journalism dying? Ads

This is not really true. Ads, especially the classifieds, were often the only way that newspapers and news organizations stayed afloat aside from public funding. What happened was the major ad-tech companies captured the revenue and the subscriber data that came from ads and viewers that would normally go to those organizations. "Real journalism" often didn't make much if not any money. "Real journalism" at least for most of the 20th century and the beginning of the 21st has always existed alongside advertising. Magazines literally survived on advertising, their entire product to readers was long form content, while, to brands, a higher quality product and more targeted subscriber base. It's pure romanticism to say that ads themselves were the problem. Ad-tech, it's monopoly control in the hands of 2-3 behemoths, is the real issue.

martokus•1h ago
Can confirm. I've spend over $1000 as a test and I see even worse results. Absolute garbage compared even to cheap garbage paid social traffic.
giancarlostoro•50m ago
If I were Mark Zuckerberg I would be salivating waiting for a call from Sam Altman.
efficax•46m ago
facebook ad targeting is also garbage. They have my entire history since 2004 and they still sometimes: think I'm a woman, think I need things for a newborn, think I'm in the market for "short men clothing" (I'm 6'0"), and so on. Sometimes I get ads for farm equipment (I am a lifelong city dweller).
necovek•9m ago
Now let's be honest: short men clothing is basically modern, "slim fit" clothing; you get that fashionable bare-ankles look with short pants.

And again, who does not want to drive a tractor in their city when they hit the pothole? Or at least to re-plow half of the city just because of the stupid buildings they put in there?

billyp-rva•47m ago
> ChatGPT recommends bidding somewhere around by £2 per click, but I bid between £0.15 and £0.07 and was still able to get traffic. However 0.46% is a shockingly low conversion rate.

I'm not an ad guru by any means, but isn't it expected that you will get junk clicks with a bid like that? Seems like a you-get-what-you-pay-for scenario.

cmiles8•19m ago
Well I think your experiment more shows that while Open AI is trying to position this as a viable marketing platform, in reality folks with marketing dollars do spend aren’t buying what they’re selling so you’re able to win at pricing far below where they really want/need the market to be.
hermitcrab•18m ago
>isn't it expected that you will get junk clicks with a bid like that?

No. Bidding low means less or no clicks, but should not affect traffic quality.

advisedwang•43m ago
> ChatGPT recommends bidding somewhere around by £2 per click, but I bid between £0.15 and £0.07 and was still able to get traffic

Did they try bidding more? It's possible they are just only getting the dregs. (In which case the ads would be expensive instead of garbage)

hermitcrab•21m ago
>Did they try bidding more?

OP here. No. The cheapest version of my PerfectTablePlan software is £19.95 + VAT, one time fee. So I would be insane to bid £2 per click.

>It's possible they are just only getting the dregs.

Bidding higher generally means more clicks, but the quality could be worse due to winning bids when your content is less relevant.

water-data-dude•41m ago
For what it's worth, the picture at the top of the article is a textbook example of very high accuracy!
jmdeon•39m ago
Wouldn't it be low accuracy but high precision?
hermitcrab•13m ago
Not if you were aiming for the center.
ok123456•40m ago
This will end in another 'adpocalypse' episode. Something like: ChatGPT generates recipes from the Anarchist Cookbook and adds Home Depot ads at the end, or inserts Amazon affiliate links for the Turner Diaries.
nikole9696•37m ago
I have no idea why anyone would click on any ad anywhere other than by accident, not knowing if the link might take them to some malware site these days.
probably_wrong•33m ago
I clicked on a couple ads once because I was desperate - I needed a cheap flight with a tight deadline and my usual search engines didn't help. I thought "if these ads deliver what they promise, I'll never book anywhere else".

Believe it or not, they did not deliver.

beezlebroxxxxxx•24m ago
I would never click on the weird ads that the supposedly "brightest minds of our generation" have delivered to me via some nonsensical middleman auction system.

But I go to some blogs (philosophy, jazz, and men's clothing) that have ads that are literally just image links. The ads are actually relevant to the content, and because I go to the blog of my own volition, that means that more often than not the ads are at least somewhat relevant and usually worth a look, like the curated ones you see in a magazine. Most other ads online are so irrelevant it's almost comical, and I have no idea why people pay to use those services. All these tech giants who make enormous sums of money from advertising are emperors with no clothes - a hard look at the quality of their product reveals it's unanimously considered garbage by the public. Their promise of delivering hyper targeted advertisement was obviously a lie predicated on market capture and becoming the only middlemen in town.

demibabs•11m ago
Sometimes (I especially notice this on the App Store, but sometimes Google), companies bid on their own name. So you'll search something like "Discord" and the first result will be an ad for Discord, and then below is the organic results which has them first anyways.
ryanmcbride•36m ago
Everything with ads is garbage. All the time. It poses no benefit to humanity even a little bit.
sssilver•31m ago
Yes.

But then, is any and all public communication about a product or a service originating from a stakeholder of said product or feature basically an ad?

WarmWash•29m ago
Except for socialising the entire internet age, giving poor people in 3rd world slums the same access to the same platforms as the richest 0.001%.

No one likes to acknowledge this, but there is an alternate timeline where only the rich kid in school has a Google account, reddit account, meta account, and almost all content is still produced by massive media conglomerates.

Ads suck and everyone hates them, but don't lose sight of what they enabled by hyperfocusing on the bad.

Zambyte•27m ago
> No one likes to acknowledge this, but there is an alternate timeline where only the rich kid in school has a Google account, reddit account, meta account.

It is not at all obvious that this would be a worse timeline to live in.

necovek•17m ago
Before Google, before Reddit, before Facebook, people all over the world, rich and poor, run their own sites and their own forums, allowing them to chat with people throughout the world.

Here in Serbia, even in the worst of times (sanctions, NATO bombing in 1999), people with internet access could participate in local and global fora, newsgroups etc. It was expensive to get a computer and internet access, but people were willing to pay to host a small community. Late in 90s and early in 2000s, there was a huge number of cheap "virtual hosting" providers, and even cheaper LAMP stack providers, where people would run their own phpBB or similar forum software. IRC I probably need not mention.

This "great equalising" has been nothing but a negative in the long run.

zenethian•31m ago
It’s wild to me that people click ads.
shawnwall•28m ago
Have you observed any humans and their decision making capabilities lately? :)
Zambyte•25m ago
Advertisements don't need to be clicked to work. Billboards clearly work for reasons other than being clicked.
nitwit005•5m ago
If you asked me to name the companies on the billboards that line the highway I commute on, I couldn't do it.

I don't think any single form of advertising is effective on the whole population.

cmiles8•24m ago
Yes, ChatGPT ads are a heaping pile of garbage. I feel sorry for anyone that’s wasted perfectly good marketing spend on such nonsense.

Outside a captive platform like Amazon, most ads are garbage and a total waste of money.

If I’m on Amazon trying to buy X and my search for X “promotes” an option to the top of the list yeah that probably influences me buying that.

But the broader online ad ecosystem is a total waste of money. Yes, I once searched a question about golf balls… no I don’t want to see “targeted ads” about golf balls following me all over the internet for the next three weeks.

Maken•22m ago
Amazon is garbage too. It keeps showing the same product that I already bought.
cmiles8•21m ago
True :-).

Probably more that from an ROI perspective that’s probably the “least worst” vs the best.

idiotsecant•15m ago
presumably if this was true advertising spend would dry up in these useless cases. The fact that it hasnt and people intensely dislike spending money without getting anything for it, I suspect it must work on someone.
EcommerceFlow•21m ago
It's interesting the ChatGPT referrals conversions match Googles. Shows once they tune a few knobs, they could have a cashcow on their hands.
jandhdhshhh•19m ago
Speaking as a consumer, I’ve always found ad targeting to be way off for Google, YouTube or even facebook. The few times I got ads I ChatGPT it was also off. It’s kinda ironic that with all that AI, they choose not to use any of that intelligence to target ads better. It makes you wonder what all those engineers are working on if not the main revenue driver.
MengerSponge•13m ago
I understand why it's hard, but it's very funny to get ads for a product I've bought for weeks/months after completing the purchase.

I looked for rain boots, and I bought a pair of rain boots. Getting ads for rain boots now isn't going to drive any purchasing, right? Meta or Alphabet get paid still, so the advertisers are getting hosed.

throwup238•8m ago
You might return them and get another brand. Even that little positive signal is often worth the wasted impressions because most of the other signals are either so weak or so competitive.
hermitcrab•6m ago
Advertisers getting hosed is pretty much standard with digital ads. Platforms and agencies tell you how amazing their adtech is, but various forms of fraud and dodgy practises seem to be rampant.
howunfortunate•10m ago
It may be that you just aren't a big spender.

My experience with performance adtech was that 95% of the energy went into targeting whales, with a focus on products that drive massive revenue (mobile gaming, gambling, crypto). It's truly insane what these whales are spending.

Sure, there's still a lot of money in the 5%, especially for a surface like ChatGPT where people are making retail decisions. But I'm not particularly shocked if they're a little behind.

jabart•12m ago
We can't use ChatGPT Ads because ChatGPT banned our company for ads for political content even though we are a CRM/Compliance tool for campaigns and the only ad we had was the autogenerated ChatGPT one they made. No chance at an appeal either. We had this issue with Bing Ads but microsoft fixed it and even personally called to apologize about the mixup.
symfrog•7m ago
I have run many ad campaigns across various platforms, and I have come away with 2 main thoughts:

1. Click fraud is rampant, and the ad platforms seem to not care in the least as long as they can drain your budget.

2. The ad bidding process is completely opaque, and no auditing seems to be done, my inclination is that there is large scale fraud here too.

snailmailman•1m ago
Is there any ad targeting that is good?

The main place I see ads is YouTube on iOS. I used to have ad targeting turned off, but I’ve since opted into all of googles tracking and ad-targeting, in attempt to curate the ads I see and improve them.

I’ve had all the Google tracking and ad-targeting turned on for close to a year now. Google still can’t even target me ads that are in English. I can’t read Chinese or Turkish or Spanish but I still get ads in those languages. Most of the ads are still scams or questionably legal products. And nearly all advertisements are blatantly targeted toward the wrong demographics. I get both “you can buy this multi million dollar hotel” and “you can use food stamps at our store”. I sometimes see advertisements for local politicians but I just as frequently see ads for politicians I cannot possibly vote for.

Maken•12m ago
Ads didn't give us Wikipedia. The vast majority of blogs and forums were not for profit. Specialised media outlets were always subscription-based. Large amounts of valuable information like research papers, technical documentation and curated databases were paywalled.

The content that ad-based websites brought was mostly clickbait designed around serving these ads.

cj•29m ago
Maybe we should look at paid media as a honeypot.

Without paid media, marketing spend gets pushed to “organic” and “content marketing” aka blog spam and other crap.

Zambyte•28m ago
Local advertisements can be nice. The advertisement section in the local newspaper was a thing people used to actually seek out, once upon a time. The real problem is the scaling of advertisement frameworks globally and algorithmically serving individuals or certain groups. It is strictly parasitic at that level.
bigyabai•19m ago
If "can be nice" is the bar of quality we hold ads to, then all advertisement passes the bar. A lot of newspaper ads were for luxury tchotchkes, cigarettes, booze or gambling. Back then, advertisements had some pretty substantial downsides, normalizing entire swathes of behavior that is today considered unhealthy and exploitative.

At most, I would call advertisement a Tragedy of the Commons. But the way publicity works, it feels like the tragedy is unavoidable; all smart advertisers will eventually transgress against normalcy.

imperio59•26m ago
The clear mark of someone who's never owned a business and had to find customers to sell their product.
howunfortunate•6m ago
I'd really like for some country to ban all ads, even if just for a while, so people could observe the quasi-experiment.

My personal bet would be the country would suffer for it quite a lot, similar to places that have banned e.g. loans, but it'd be very cool if I were wrong!

c7b•9m ago
I've been down the same thought path, but my conclusion was that it's a fallacy. Google and Facebook make literal billions with data-driven ad targeting. I think the more apt conclusion is that what would you and I would consider a good or at least tolerable ad experience is not what maximizes their profits.

I vaguely remember a blog post by a former FAANG engineer on HN lamenting the tyranny of designing for the marginal user (ie the one that gets swayed by an epsilon of optimization). That marginal user has incredibly short attention span, makes incredibly dumb decisions and overall has the cognitive capacity of some plankton.