Hacker News .hnnew | past | comments | ask | show | jobs | submit | charcircuit's commentslogin

No it doesn't. Unlike Brave, Firefox needs an extension to block ads just like Chrome.

Yes, though it has the most powerful and customizable adblocker available.

I've found Brave's built in ad blocking to be good enough on its own.

What ads are slipping through?

The video is privated now, but the timelapse is weird. Sometimes it skips only seconds before the next screenshot and sometimes it skips probably hours forward.

>I’ve seen evidence they lose money on heavy users.

Where?


There are tons of blog posts where folks work out the API cost of their usage and find it well above subscription cost.

That doesn't mean the company is losing money in aggregate on these subscriptions. Buffets are still in business even though some people gorge themselves silly at them. The incremental cost may exceed the incremental revenue for a particular person or minority group, but that's not how these businesses measure profitability.

There is a difference between making less profit and losing money. Comparing to API cost can only show the former.

>During early testing, Stripe reported that Fable 5 compressed months of engineering into days. In a 50-million-line Ruby codebase, the model performed a codebase-wide migration in a day that would otherwise have taken a whole team over two months by hand.

Who is refactoring by hand? This comparison is not relevant in 2026.


Embedded systems can make network calls to powerful, GPU equipped servers.

Sure. Claude does that. "Cogitated for 1m 50s" doesn't work for real-time applications.

You can submit many queries in parallel to increase throughout. Smaller models and faster hardware can reduce the time per query too.

None of that gets you the 100ms response time the parent poster talked about, for something like "who is at my doorbell?" real-time uses.

Ok. Claude will not work for this use case because none of the sample data (weirdly blurry ID images) is in the training data.

They really shouldn't, though.

It can offer a ton of user value. There is a whole industry built upon this idea, Internet of Things.

IoT wasn't not built on "send all the data off to a hosted GenAI". It predated them by quite a few years.

The GPUs were doing video transcoding instead of GenAI.

You can run OpenCV on a GPU-less Raspberry Pi or other IoT device just fine.

And most IoT devices aren't doing video transcoding at all. You're making some very odd assertions in this thread.


>And most IoT devices aren't doing video transcoding at all.

The data gets streamed to the cloud where servers with GPUs transcode it. I'm pointing out that IoT devices historically have reached out to servers with GPUs even before GenAI.


Most IoT devices have no camera and communicate with servers that have no need for a GPU at all.

>for very limited benefit

Due to how simple they are to work with they will become popular. Compare NLP before and after GPT-3. GPT-3 majorly brought down the complexity and skill needed for doing NLP tasks even if traditional NLP is much much faster. Ultimately ease of development will win out and the industry will work towards optimizing running such LLMs to make it cheap enough to run.


Why not use those extra fiber optic cables as part of the actual internet? You can immediately start making money off of it due to people getting legitimate value from it. If other lines were to be cut or have issues your lines would act as a backup. And yes if things got overloaded it would act slower. It seems unoptimal to make a completely separate network.

There is ton of utility. I use it all the time to study, to look up what's happening in the world, to understand the context behind what others are saying, cooking recipes, and much more. Considering LLMs have access to tools for searching the internet they have a superset of the capabilities of Google and consumers got a lot of value from Google. In fact from putting ads on the search results Google has made billions of dollars from such consumers getting value from their service.

When you ask it to give you a digest of current events or as a study aid how are you ensuring that what your reading is a valid representation of the source material? Has it never given you false information?

Not OP but, anyway, AI output should be treated like any other source material.

I study from reputable sources every day and never cease to be amazed by how many errors or misconceptions they have. Peer-reviewed articles, books from renowned scholars, news from major publications… regardless of the source, false information and contradictions accumulate. I’d wager that AI, besides helping me uncover these issues in the literature, has had a lower error rate than most of the materials that I read on a daily basis.


I can tell by my intuition. If I'm that interested in something I can dig deeper or just give up in pursing the truth.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: