Brave Search launches own image and video search (opens in new tab)

I accidentally found a good test string for image search a few months ago:

Bamboo sign

Give it a try on Google images. You'll see that nearly all the results are x-rays of people with ankylosing spondylitis, a form of which is commonly referred to as "bamboo spine".

I tested out brave search - it correctly shows 90% signs however it does also show a few spines.

Google still shows incorrect results months later. It's by far the worst of all the search engines in the list for this simple and obvious search.

Nice!

I actually like Brave here for my test better than Google. I typed in a few cities, just wanting to see the skylines and such.

Brave gives me good photos, some stock photos, etc. Google gives me pictures from recent news articles, which isn't overly helpful IMO.

Brave is the one that censors less, from all those. Specially doesn't censor for political motives that I'm aware of.

That already makes it worth of support.

But Google having become so bad of late has made switching quite easy, even if brave is not getting better super fast, Google unfortunately is getting worse and making up for it.

You missed out Mojeek. For an independent take: https://seirdy.one/posts/2021/03/10/search-engines-with-own-...

Shouldn't Kagi (https://kagi.com) also be on that list?

One more, which is self-hosted, peer-to-peer and FLOSS: https://yacy.net

does searx count as independent?

ahrefs

c.f. yep.com

Hmm, I don't know, it doesn't seem obvious to me that it is unethical to disobey the publisher wishes.

If you post something to the open web, what's it to you who reads it and how? You can block some IPs but that's about it.

I don't know if Brave has a knowledge graph - if they do, I would understand objecting if they filled it in with “stolen” content. But I don't see what's the problem with search.

By the way, isn't everyone's favourite archive.is doing the same thing?

I have no strong opinion on this, curious to hear counter arguments.

This make me want to use Brave search now. When I use a tool I expect it that it serves me, not the material it provides.

> A publisher should have full control to discriminate which search engine indexes the website's content

If you want someone to not see what you publish block him yourself. Also why would you want to do that? Do you want google to own the web or something?

Let's say I pay for Kagi. Kagi is a tool that I'm using to avoid doing hard work manually. With relatively few exceptions, I can probably accomplish what I use a search engine for manually, but with much more time and effort. So I'm paying for a tool to assist me with my use of the web. A "user agent", you might even say.

It simply doesn't sound right to say which tool a user can use. It's literally the same as arguing that you should be able to block Firefox from accessing your website and it's Mozilla's fault that they don't respect your wishes as a webmaster to block Firefox exclusively. Or that a VPN doesn't publish its IP addresses so that you can block it. Or a screen reader that processes the text to speech in a way that you disagree with.

Philosophically it seems intuitive to say "I should be able to block a third party that is abusing my site" but it's ignoring the broader context of what "open web" and "net neutrality" actually mean.

I run a service for podcasters. There are podcast apps and directories that either ignorantly make unnecessary requests for content or have software bugs that cause redownloads. I could trivially block them, but I don't because doing so penalizes the end user who is ultimately innocent, rather than the badly behaved service operator. The better solution is primitives like rate limiting, which I use liberally. Plus, blocking anyone literally has a direct effect of incentivizing centralization on Apple, Spotify, etc. and making the state of open tech in podcasting even worse.

> the Brave search API allows you (for an extra fee) to get the content with a "license" to use the content for AI training? Who allowed them the right to distribute the content that way?

I don't think there's any court at this point that would back you up that freely published content annotated with full provenance cannot be scraped and published for a fee. Services like this have existed for decades. If you don't want your content scraped, put it behind a login. Especially considering this only applies when you allow other search engines and if you think Google and Bing aren't using your content to train AI, you're off your rocker.

They are right and you are wrong. If some web page is publicly available, it should be indexed. Scraping neutrality, please.

> The Robots Exclusion Protocol is a mechanism for publishers to discriminate between what users and crawlers are allowed to access, and discriminate between different crawlers (for example, allow Bingbot to crawl but not Googlebot).

To me as a search engine end user, this kind of behavior is undesirable. Why would I want a website to selectively degrade my experience because of my choice in search engine or browser?

Brings back horrible flashbacks of “this website is only compatible with IE6”.

Curious why cannot selectively block using IP address instead of user-agent string. According to HTTP specification, UA is not a required header. There is certainly no technical requirement for it in order to process HTTP requests. Of course, any website could block requests that lack a UA header. I never send one and it's relatively rare IME to see a site require it, but it's certainly possible.

You should check these assumptions, Mozilla has been hard at work enshittifying their entire portfolio. Instead of giving the public features they actually want (the most secure, performant, and predictable web browser), the current CEO has directed the focus towards revenue-generating features.

My god, there is so much telemetry in FF now, and it's tricky to hunt down all the about:configs to disable it. Not friendly or privacy conscious at all. Do you really want Mozilla to get pinged with your IP address every time your browser process starts and exits? Yuck!

Quick FF enshittification example from 2 months ago:

Alert HN: Mozilla puts advertising into Firefox AGAIN

https://news.ycombinator.com/item?id=36351322 (48 comments)

Mozilla stops Firefox fullscreen VPN ads after user outrage

https://news.ycombinator.com/item?id=36085642 (220 comments)

> I'm always staying away from Brave because I've been confronted so many times with bait-and-switch tactics that I have the feeling that one day they will move away from being good and monetize all the collected data, even though they don't collect data.

The thing is that it will always get worse. Every company needs to grow in order to stay alive and eventually quality and your user experience will suffer because of this, but here's the thing: Always choose the upstart competition but just be prepared to jump to the next up-and-comer after that. For me, I found DuckDuckGo getting worse over time (probably not on purpose, just spam) and somehow Brave is better so I'm sticking with that, but as soon as Brave decides to fuck me (and they will!) then I'll be jumping to whoever the new underdog is at that time.

Hi, I'm David and I've worked along other engineers at Brave on this! Thanks for your feedback, it would certainly be a nice addition, although we may want to focus a bit more on quality first. Thanks!

I’ve been using tineye.com

Bing visual search is great. I use it in cryptic hunts a bunch. Would of course prefer a non-MS offering though, I hope one comes along that's equally good.

Bing has its own and it's not bad.

Usually google+tineye+bing+yandex will yield some useful results

Same here. Much like Firefox, I think it's important to use a search index that isn't tied to the larger players (Bing and Google in this case). I tried Brave a couple of months ago[1], I found the results better than Bing, but without image search it wasn't usable. Now I can give it another go.

[1]: https://jahed.dev/2023/07/01/trying-brave-search/

i have been using brave search for a while now and i was surprised and am very satisfied with the search results. missing image and video search was a bit annoying mostly in that it linked to google and bing but not any other search engines. but i just remembered to switch to where i wanted to do image search instead. it sometimes meant that i had to go back to retype my search query, but i'd rather have a good text search than be bothered by that. in most cases i'd know ahead of time if i wanted images so it was easy to pick the right search engine.

On the topic of "weighting", Brave Goggles allow you to decide the weighting/ranking yourself, and is a great and under-used feature. The UI is a bit lacking (you have to create a goggle as a github gist/repo, publish it by submitting it to them, then bookmark an unweildy URL that's the link to your custom search engine), but the goggle syntax is pretty expressive and easy to use.

this is exactly why i switched from ddg to brave

I'm not sure if they can without rewriting the whole thing, the original (WebKit/KHTML) code base being GPL.

On the other hand, the Google lawyers seem to have found an excuse to link some proprietary code into Chome (that's not part of Chromium). Does anybody know what that excuse is, and if it provides a loophole large enough to close off Chrome development?

I see they say "if a domain or page is not crawlable by any search engine (it has a noindex tag), or if it is not crawlable by googlebot, then Brave Search’s bot will not crawl it either."

1: https://brave.com/search/api/

We don't collect any data, so there's nothing to "monetize". We bring ad matching and blind confirmations to the browser, no need for collection.

You have to realize data is worth less over time -- a lot less. So the issue would be whether users stick if (heaven forbid) a change of control left Brave in the hands of unethical owners. By design, open source, network sniffing, and other auditors would catch wise and flame such a corrupt Brave into the ground. That's the best we can do here. There is no safety property ("X" holds for all future program states) enforceable on a company.