Nostr Web Client

Ok. Been meaning to publish this for a little while now.

I think reviews are an extremely underrated / under explored design space for application that want to leverage Nostr (in particular, the social graph).

In this essay, the first of a series I want to publish on nostr:npub1w0rthyjyp2f5gful0gm2500pwyxfrx93a85289xdz0sd6hyef33sh2cu4x, I break down:

- how reviews drive commerce

- despite their being broken / fake / easily gamed

- how the internet originally won

- what’s needed to make the internet functional again (upgraded stack)

- how this upgraded stack makes a whole new design spaces possible

- how ONE of the many things we’ll fix is reviews

- how we’re going about it with a project we’re calling Satlantis

https://highlighter.com/npub1dtgg8yk3h23ldlm6jsy79tz723p4sun9mz62tqwxqe7c363szkzqm8up6m/Reviewing-Reviews-lpinxc/

nostr:npub18lzls4f6h46n43revlzvg6x06z8geww7uudhncfdttdtypduqnfsagugm3 - as promised. Put it up on highlighter.

nostr:npub1sg6plzptd64u62a878hep2kev88swjh3tw00gjsfl8f237lmu63q0uf63m - this is what I was alluding to last week. That being said, I want to dig deeper on your particular comment, and I’ll do that in the next essay.

nostr:npub1sfhflz2msx45rfzjyf5tyj0x35pv4qtq3hh4v2jf8nhrtl79cavsl2ymqt - Thankyou for your early feedback 🙏🏽

Reply to this note

Please Login to reply.

Discussion

IntuitiveGuy☯️ 1y ago

interesting.. #nostr social graph could indeed create a different level of value in reviews..

i would rather buy a product that has 1 good review signed by a Nostr user that I know and is part of my social graph and feed than a similar product with maximum rate reviews made by thousands unknown and maybe fake users

Leo Wandersleb 1y ago 💬 1

The problem is that for an informed opinion, your social graph should at least rate a few of these things and assuming Dunbar's Number as the number of relevant follows you might reasonably have, if you reduce that by shared interests you quickly fail to find any relevant opinions. What are the chances that any of them has been to that other country, to that other town and visited any of those 10% of still available hotel rooms you can find for your travel dates? Or even more generic items like headphones ... how many of them have tested headphones that are still on the market in your town ... and taken the time to write reviews?

Google has the social graph and uses it on the few occasions there is a match but it's few and far between.

Svetski 1y ago

This is a good point. I guess a potential solution to this is concentration around particular verticals ?

brugeman 1y ago 💬 1

It's a very good point. Those billions of reviews globally, if you exclude fakes and then divide by the numbers of products out there you get maybe 100 reviews per popular product (median number of reviews for all products is definitely zero). How many of those 100 reviews would be in my wot circle? If the 100 people I follow posted 100 reviews each, I would only be considering maybe 0.01% of all popular products. That doesn't seem good.

The solution could be a wider wot circle, and start from niche, and more reviews. If I knew my reviews wouldn't be lost in some corporate database, I might have posted more of them.

Also it will probably vary based on physical proximity to my contract list members. They would probably review a lot of local products, but not much in Dubai or on Alibaba. So maybe travel is not the best place to start.

Svetski 1y ago

Interesting. Appreciate the feedback. Much to ponder here

brugeman 1y ago

Also wrt 'headphones' example. Instead of looking through my wot circle for reviews, I could post 'please help me choose headphones' message that could propagate through my wot circle until some experts see it. They could reply with 'here is a one liner big advice, and here are a couple links for you to read, and if you want to hop on a call - schedule here and it will be 5000 sats'. This might turn out as a better solution to 'help me buy headphones' problem than trying to fix fake reviews or increase the review density in my wot.

hodlbod 1y ago

Also vectorized trust can help find the right people from your web of trust for a given product.

brugeman 1y ago

Can you please tell me mo about vectorized trust? Maybe a link to read?

Leo Wandersleb 1y ago

Working on nostr based feedback myself. Check out these long form articles

https://habla.news/t/WalletScrutiny

They are all reviews of Bitcoin wallets and I share the enthusiasm of your summary above. Looking forward to having time to read the article.

Leo Wandersleb 1y ago

Is it meant to be a wall of text?

Jason Hodlers 🪢 1y ago

No, it should look something like this:

https://nostr.at/naddr1qqv9yetkd9jhw6twvuk4yetkd9jhwuedd3cxjmncvvpzq6ksswfdrw4r7mlh49qfu2k9u4zrtpextk955kquvpna3r4rq9vyqvzqqqr4guc7ry3k

arkinox 1y ago 💬 1

💯

I think you guys are using QTS for Satlantis, right?

https://habla.news/u/arkinox@arkinox.tech/DLAfzJJpQDS4vj3wSleum

nostr:npub149p5act9a5qm9p47elp8w8h3wpwn2d7s2xecw2ygnrxqp4wgsklq9g722q

Svetski 1y ago

That’s the plan, yes

vinney...axkl 1y ago

Weird, this post keeps freezing the highlighter UI for me

Jason Hodlers 🪢 1y ago

Same here. 🫤

Jason Hodlers 🪢 1y ago

This works for me, though:

https://nostr.at/naddr1qqv9yetkd9jhw6twvuk4yetkd9jhwuedd3cxjmncvvpzq6ksswfdrw4r7mlh49qfu2k9u4zrtpextk955kquvpna3r4rq9vyqvzqqqr4guc7ry3k

Kip 1y ago

really looking forward to this!

What do you think about adjacent spaces, for example performance track records of money managers? I think there are a lot of similarities...easily gamed, don't transfer with the person, not verified etc.

I have a vague inkling a nostr like architecture could fix this

Bill Cypher 1y ago

If I could look at reviews and see the reviews from my follows and followers separated out of the crowd that would go a long way to restoring my trust in their having a utility in predicting product quality.

If I could zap a useful review I would. If I knew my review wouldn't get lost in a sea of AI spam and might get zapped, I would write more reviews.

The utility difference here is big enough that this review system being on one site and not another would change where I bought things in some cases.

nostr:nevent1qqstz7lyhcqkasywu8z3z7g5umsndwjdjvh59fw663q3pykzt8vl73spzpmhxue69uhkummnw3ezuamfdejsygr26zpe95d650m07755p832chj5gdv8yewckjjcr3sx0kyw5vq4sspsgqqqqqqs4hj5f5

Max 1y ago 💬 4

Great essay! Nostr is definitely the way to go for a reviews site. Probably starting with Bitcoin adjacent reviews.

I've got a piece coming soon that considers some other ways Nostr identity/WoT might be useful. It feels like these use cases are about to explode...

nostr:note1k9a7f0spdmqgacw9z9u3fehpx6aymye0g2ja44zpzzfvykwelarqv4drqc

miljan 1y ago 💬 3

Nostr fixes reviews & ratings.

EVERYTHING WILL RUN ON NOSTR. NOBODY IS BULLISH ENOUGH.

nostr:note1k9a7f0spdmqgacw9z9u3fehpx6aymye0g2ja44zpzzfvykwelarqv4drqc

wavy. 1y ago 💬 1

You have a source for your review data ? Seems wildly outrageous numbers. Maybe im wrong but i doubt many consumers who buy items like Supreme or Palace ever needed a review… Reviews are definitely valuable but why couldnt i just create a fake acct and pump my own products the same way ? Nostr fixes that because of how exactly ?

Svetski 1y ago

I couldn’t link the sources to that essay. It kept bugging out the highlighter post.

Keep an eye out on my medium or Substack for that.

Svetski 1y ago

Repoasting:

https://highlighter.com/svetski/Reviewing-Reviews-lpinxc/

Bit McCoin ⚡ 1y ago

Why do we need to use the social graph if we can just use zap counts directly as a signal of relevance?

diasporic 1y ago

The zapvertising astroturfing risk?

It could end up as yet another pay-to-play economy, except advertisers really ought to be paying the end-users for their attention, rather than paying themselves to climb the relevancy ladder.

Bit McCoin ⚡ 1y ago

Locking up liquidity has an opportunity cost. Only worth it if users find content engaging enough to tip.

diasporic 1y ago

Alex: Thanks for the thoughtful presentation.

Do you think it's possible to calculate transitive trust/relevance based on each principal's social graph?

Would it be desirable to have a dampening function on such a transversal search, a la IP routing protocols' TTLs. Otherwise you might end up with social graph loops.

Also, I may care less about the judgement of a friend of a friend of a friend than I do about a direct friend.

When I think about webs of trust, I immediately think about how unwieldy the PGP user trust survey is. Also, I think I'd feel a bit uncomfortable about broadcasting my evaluation of my social graph connections. So if there are any gradations between "connected" and "not connected", it seems like that ought to be private information either stored locally for each client/user, or else self-signed/encrypted data if published for cross-Nostr-client portability.

OK, one last blurb re: open social graphs. I think Tantek Çelik's microFormats, hCard, XFN, and FOAF concepts might be relevant, although they all seem to lean towards the too-much-information/self-compiled-dossier side of social graphs. See http://gmpg.org/xfn/and/#idconsolidation for some links.

Thanks again!

3eppo 1y ago

great post!

one major issue i see in reviews is that a lot of people have no incentive to post a honest review at all. either they post a 1star rant review or they post a 5star brain-afk review for getting a coupon.

in our company we force our technical procurments team to write one good and one bad thing about every supplier used in a project. and also give an overall performance mark (1best to 5worst). it doesn't matter if we had 1 or 100 orders for this project at that specific supplier. we need this for our iso certification.

nevertheless the comments seem to have no real value at all, as they either write "-" or a wall of text no one reads. the grading depends actually on the people. there is one group who almost everytime gave a non meaningful grade (3,3,3,3,3) and the other group gave extreme grades (1,5,5,1).

however we change the process so that the team has to grade during the process (after the order, after the incoming goods inspection, after the component is installed) simply by giving a 1 to 5. i know it doesnt solve the problem of subjectivity but it gets us closer.

tldr.: think about: incentives reviews, directly integrate them into the process, normalize reviews

Bitcoin Boomer 1y ago

Great article! Thank you.

Nick Slaney 1y ago

I’m here for it, the problem is clear, how can I trust a reviewer? How do I know they are really qualified to have an opinion here? Have they eaten a lot of Indian food or did they grow up on chicken nuggets and tried something spicy for the first time in their life and that’s why they left a bad review?

I’m not sure if reputation can be solved with tech. It is deeply social and very hard to create a system that doesn’t suffer from the first order effects you list here. (The next meta game of reviewing the reviewers).

It’s like the ultimate solve for online everything though (partially why I worry it is not achievable / an endless rabbit hole).

The govt relays reputation (imperfectly) through licensing. Uber took that system and brought a market place with centralized but crowd sourced reputation provided by them to great effect. If you can design an open system that provides generalizable reputation for any service, you unlock a truly open online marketplace for any service.

All that being said, excited to see what you come up with!

Satstack 𓆀⚡︎ 1y ago

🤙🧡

⚡ Dee Kay ⚡🇸🇪🇬🇧🇨🇿🇧🇷🇦🇹 1y ago

Spelling?