5 ms·
Good stuff, however comparisons to duck and Google for first billion users is misleading since internet use has exploded exponentially since last decade. You’d
by endisneigh 4y ago
Good stuff, however comparisons to duck and Google for first billion users is misleading since internet use has exploded exponentially since last decade. You’d expect it to be easier to get more nominal use.
Also, as much as I dislike ads I think web3/crypto grifting is worse. I feel ambivalent =\.
I’m curious to read how Goggles is architected technically. Allowing anyone to create their own defacto index seems intractable with Google level performance.
- vlunkr 4y ago> internet use has exploded exponentially since last decade Also the number of bots has exploded, that must account for some of the searches.
- ryan29 4y ago> Also, as much as I dislike ads I think web3/crypto grifting is worse. I feel ambivalent =\. I super briefly skimmed the crypto (BAT) token stuff. I think the general idea of earning tokens for searching and browsing is neat, but the way they're trying to use it directly as a microtransaction currency isn't super innovative. I don't like crypto in general, but I think the "proof of idle capacity" tokens have stumbled onto something that could be valuable for online reputation because they invert economies of scale in a way. Take something like Chia (XCH) as an example. As a typical user I can allocate some idle storage space to mining and, as long as I only mine while I'm browsing the internet, the incremental cost to me is basically $0. Compare that to someone building a rig specifically for mining. They have costs, I don't. Assume I can opt in by plotting some proofs on idle storage and mining into a pool where rewards are split up and some of the rewards (aka tokens) can only be used for liking (ranking up) or disliking (ranking down) websites. Ex: 20% of your rewards go to the pool operator (search index), 5% can be donated directly to sites or creators you like, and 75% go to ranking (search reputation) which "burns" them. Adjust the ratios as needed. A million normal users with 10GB of spare storage cancels out 10PB of storage from bad actors trying to game the system. IE: What's free for the masses is expensive to game and the more users hate your site, the more expensive it is for you to game the results. If you never give users "real" money (aka redeemable tokens) as rewards there's not much incentive to buy storage just to "earn" crypto currency, so hopefully you'd keep it limited to using truly idle resources only. The search index (ex: Brave) could use the tokens as a currency for selling ads (ie: keywords) and that could be weighted by the reputation score of the site you're driving traffic to. In other words, the effectiveness of your ad buys is impacted by your existing reputation and if you have a site that everyone hates you'll eventually get to the point where ads for your keywords cost significantly more than competitors that are well liked. Now everyone on Hacker News can tell me why that's a dumb idea :-) Lol.
- dubswithus 4y agoThere have discussions of requiring staking BAT for advertising on Brave search. Probably won't happen because BAT is meant to be a utility token and this could piss off the regulators. > I think the general idea of earning tokens for searching and browsing is neat, but the way they're trying to use it directly as a microtransaction currency isn't super innovative. Micropayments have never worked before. There's hope in the future if the network effect is strong we'll be able to pay a little bit of BAT for one NY Times article, etc.
- philjohn 4y agoWhat was humorous was a friend of mine who is super privacy conscious, pihole, cookie blocking, javascript blocking etc. etc. They complained that the ads they're getting aren't interesting and are just in general crap. Yeah, that's what the web was like before advertisers could go "hey, if you have people interested in X show them this ad".
- rglullis 4y agoTo improve the personalization, they "just" need more diverse advertisers and work on their in-device algorithm. You don't need to give your data to Google or Facebook to get better ads.
- philjohn 4y agoThe problem with in device ads is that they break down when you: a) don't own the device b) the device is very low performance, or has no local storage c) accessing via the web d) don't have enough data allowance for the much larger overhead to download Now, those problems will go away eventually, but to say that on-device ad targeting is a panacea for all ignores the majority of internet users around the world today.
- rglullis 4y agoCalling it "in-device" is just shorthand for "self-hosted". You can have something that works like pi-hole and also manages ad inventory, runs the algorithm for ad-matching, checks if there is any kind of notification that you might be interested, etc, etc... all without having your personal data leaving your home.
- dubswithus 4y agoYou think Brave paying users to view ads and doing machine learning locally is worse than Google tracking you and showing you ads?
- endisneigh 4y agoYou think Google giving small businesses a chance to reach their 2 billion plus users is worse than giving people volatile internet points? Where do you think the money comes from? At least Bing gave people legit cash.
- dubswithus 4y agoThe money comes from advertisers. The advertisers pay Brave and then Brave buys BAT off the open market to distribute to users.
- endisneigh 4y agoA similar logic can be applied to Google - money comes from advertisers which allows users to use search, gmail, maps, etc for free.
- dubswithus 4y agoI prefer to be paid with a revenue share. It's a better deal for me.
- endisneigh 4y agoIt’s good to have options, indeed.
- cbm-vic-20 4y ago"I reject your ads and substitute them with my own!"
- 4y ago