Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Ask HN: What happened to Web3 startups?
67 points by Freddie111 on July 9, 2023 | hide | past | favorite | 99 comments
I guess most of them still have some runway left with the VC or ICO money they raised. How is it to work at such a place after the hype and after the sentiment turned negative?


See Molly White's "Web 3 is Going Just Great"[1]

She keeps a tally of all the scams. The total is now over $67 billion.

[1] https://web3isgoinggreat.com/


> in August 2021 they were exploited for an (at the time) record-setting $611 million.

> Now, it's happened again, and some reports are throwing around even more massive numbers like $42 billion. In reality, the exploiters were able to mint massive quantities of tokens on multiple networks, with their wallet balances showing numbers in the billions. However, complete lack of liquidity for these tokens meant their "billions" are worth substantially less.

> According to crypto research firm Beosin, the attackers have so far cashed out around 5,196 ETH (~$10.1 million) in liquid assets. Poly Network suspended services shortly after the attack.

Unbelievable hahaha


Nearly all of these are blockchain/NFT stuff and not decentralised internet infrastructure. There's probably more money stolen from physical highway construction than Web3 infra.


> Nearly all of these are blockchain/NFT stuff and not decentralised internet infrastructure.

web3 is 100% blockchain/NFT stuff (aka scams and ponzi schemes) and 0% decentralised internet infrastructure.

Also, internet infra is already decentralised.


My friends in Web3 are now AI experts building chatbots


I think everyone knew they’d move onto the next scam


At least applications built upon the “we call the OpenAI api” product strategy are less scammy than Web3 felt.


Never fear, we’ll get there


There are definitely a bunch of outright scammers already operating in that space. A bunch of the “AI-powered” browser extensions are just malware, for example. https://www.pcmag.com/news/beware-ai-scams-meta-blocks-1000-...


Exactly, the AI chatbots is the perfect technology for autoscamming people.


Hey, you know… that’s a use case! Better than Web3’s common use case of… ?

</sarcasm>


I think scam is perhaps not quite right. I see AI as the evolution beyond social media algorithms. Now it will feel even more human resulting in the manipulation being even more subtle and effective over time. Being more human like may also help people feel like sharing more information with it. Some may even think it is their friend. I think there is a lot more psychology in play with AI whereas Web3 mostly plays on ones greed and facilitates wash trading and faux investments into unregistered unlicensed securities. So for AI I believe a better description could be manipulation framework.


Comparing web3 scams to this new wave of AI is unironically ignorant and delusional.

For one small example I literally get daily value from Copilot and happily pay for it, and there is much more beyond this. I've never once gotten meaningful value from web3.


These people migrated to AI because their goal is to get rich quick. Whether or not it's a scam is orthogonal. It's the same people got excited about metaverse, producing online courses, and weed startups. If they aren't in tech they're doing supplements and drop shipping.

Unrelated, but I don't think AI is a scam.


That's not the point. Yes, there are many legit startups, exploring the very real value add that AI has to offer.

But the people that previously sold scammy web3 stuff did move on to peddle scammy ai stuff.


I don’t think Copilot and whatever fresh hell the popularised version of AI is are comparable.


like 90%


The only cash flows into the ecosystem have been from ICOs. The SEC is cracking down on such activities.

Besides investment cash flow, I don't think there have been any revenue generating ventures that cater to customers outside the Web3 ecosystem.

The true web3 players will be traditional finance companies upgrading their settlement systems.

I am curious how people will answer your second question.


> The true web3 players will be traditional finance companies upgrading their settlement systems.

That happens in the US this month, but it has nothing to do with crypto. The Federal Reserve System is launching FedNow, a 24/7 instant settlement system which obsoletes middlemen such as Zelle, Paypal, and Venmo. This is like Pix in Brazil, SEPA in the EU, WePay in China, etc. - a fast payment system run by banks and settled through central banks.

It's much like FedWire, but faster, cheaper, and for smaller amounts. You buy houses with FedWire; you buy dinner with FedNow. It's an inter-bank system, like Visa; you have to have an account with a bank to use it. Chase and Wells Fargo are on board, but Bank of America is not.

This will bring the US up to date with the other countries that have national payment systems.

Anybody in the US involved in moving money around needs to get up to speed on FedNow. Fast.

[1] https://www.frbservices.org/financial-services/fednow


one of the really nice things about SEPA is the 0-fee instant transfer (ime 45-60 seconds). Will FedNow be able to do the same thing?


The Fed will charge banks a fee of $0.045 per transaction.[1]

Transactions take a few seconds. They must finish within 20 seconds or they time out, and they're expected to take much less time. Here's a technical overview.[2]

The bank to Fed to bank API uses IBM MQ and ISO 20022. These are a bit dated but widely used in the financial sector. Down at the bottom, it's XML.

[1] https://www.frbservices.org/resources/fees/fednow-2023

[2] https://explore.fednow.org/resources/technical-overview-guid...


Thank you for posting the actual technical pdf, that was interesting reading

> Certificate-Based Authentication

Heh, so I can look forward to the biennial banking outage as they forget to renew their certs :-D

But in seriousness, I have some experience with x509 auth from the early days of kubernetes using it and unless the system has some robust CRL mechanisms I look forward to knowing how such a scheme will deauthorize users

Interestingly, they also seem to be doing belt-and-suspenders since the messages themselves are signed with the private key of the institution, which appears to be registered out of band from the system itself


Dated?! By banking tech standards, ISO 20022 is futuristic magic, a mere decade old. A lot of banks (particularly in the US) would still be using ISO 15022 or even older standards.


Crypto is 1-2 seconds for about 0.0025 USD worldwide right now.


Which crypto? Where can I pay with it? What if I don't want to have a hardware wallet everywhere with me? Can I cancel transactions and get my money back? Does it use more energy per year than Argentina?


Good questions.

Solana is the largest of the cheap/fast networks - I’ve worked in Solana for a year - but Aptos and Sui also have solid tech.

You use Safari on your iPhone and unlock your wallet with FaceID.

Right now physical stores that take USDC are generally luxury goods marketplaces, physical art galleries, and PC hardware marketplaces.

No you can’t cancel transactions after you’ve made them, however there are escrow services that provide similar facilities for a small fee.

No it’s proof of stake so the environment is fine.


This sounds promising, I'll get back to you in 10 years.


A completely reasonable reply. There’s a lot of good tech around, it’s definitely better than HN’s 2017-era crypto knowledge, but it’s not evenly distributed at all. I agree the proof will be in a decade.

Personally I jumped into Linux when it was a joke and everyone used Solaris, Python when it was a joke and everyone used Perl, and virtualisation when nobody would ever use it outside testing. I know I’m early in crypto. I hope it will work out but there’s no guarantees on anything.


> There’s a lot of good tech around

You mean Solana which has to take its network offline every two weeks or so [1] because it can't cope with yet another error/ddos/misconfiguration/what not?

[1] exaggeration.


Oops, during your transfer the value of the currency is now 80% of what it was!



Really? What crypto exactly and how?


yea well i pay people in money not crypto


Sounds like ACH


It is like ACH, but it settles in seconds, rather than in hours.

The current "same day" ACH schedule is that payments that make into the day's batch must be received at the Fed by 4:45 p.m. ET, a target file distribution time of 5:30 p.m. ET, and a settlement time of 6:00 p.m. ET. It's all batch files.

Anything that re-bills periodically will remain on ACH. ACH handles requests to debit someone else's account with ACH, and they can undo that later. That runs on a 3-day cycle. FedNow has to be initiated from the money sender end.


Just for the regular Joe.


> The Federal Reserve System is launching FedNow, a 24/7 instant settlement system which obsoletes middlemen such as Zelle, Paypal, and Venmo

So exactly like crypto except US specific.


Yeah, except for the fact that regular people will actually be able to use it to buy potatoes using a real currency, not a "currency" that's more volatile than tulip bulbs were in the 1600s.


i’m not sure if you’ve use crypto before, but people buy things in USDC all the time.


I'm new at this, I need a bag of potatoes. Where can I buy them with USDC?


Where can I buy them with Fednow?


Right now, probably nowhere, it's just being introduced.

Afterwards, if it's anything like SEPA, aside from my joke "bag of potatoes" comment, everywhere. In Europe you can make SEPA or SEPA backed transfers almost everywhere.


I live in Europe and have never seen a single potato retailer advertise SEPA. Electronic payments are all Visa and Mastercard.


> The true web3 players will be traditional finance companies upgrading their settlement systems.

1000%. The one place where I think blockchain and crypto can have real utility is from large, legacy financial institutions upgrading their very, very, very antiquated backend settlement technology and processes.

Of course, backend enterprise tech is pretty "boring", and the transition will take many years/decades given that our existing financial system currently depends on these old settlement systems that are based on manual, file based-processes from the first half of the last century. This is not the kind of technology transition likely to attract celebrity endorsements or bubblicious "crypto towns" like Miami and Austin.


> The one place where I think blockchain and crypto can have real utility is from large, legacy financial institutions upgrading their very, very, very antiquated backend settlement technology and processes.

And what is the actual utility beyond just the (usually unnecessary) upgrade?


You can see the revenue generated by Ethereum here:

https://defillama.com/fees/ethereum

The reason why protocols are able to generate cash flow is precisely because there is demand for blockspace by customers.


I work fulltime at a Web3 startup & have been for almost 5 years, and we're building a non-custodial smart contract wallet + everything else you need for web3 SDK & a trading card game using web3 tech. the b2b stuff is going great, tons and tons and tons of integrators and devs using our SDKs every day. the sentiment in the gaming world has always been pretty negative towards web3, so nothing feels different now ;)

large multinational brands are dropping nft collabs often right now - adidas, tons of high fashion, etc. & many other other big brands are working on them behind closed doors

i don't think it's going anywhere, and most people are just building out tech and infra


Had to scroll 75% of the comments section to find an actual on topic answer and not just inane cryptoskeptic rants


The person you responded to is a bit of a skeptic. They wrote "i don't think it's going anywhere"


Not a native speaker, but I read that as it is here to stay, not that it is pointless.


Yes, that's quite an interesting little quirk of linguistics. It could mean 2 complete opposites: that either Web3 will be successful, or that it will be unsuccessful. Maybe it'll *get off the ground", or maybe not. Maybe it won't "go anywhere," who knows!


> the b2b stuff is going great, tons and tons and tons of integrators and devs using our SDKs every day.

But how much revenue is this pulling in? Are you profitable after 5 years?


Oh, no, someone has opened a hole in the space time continuum to 2021! (Based on the Adidas thing, anyway)


I mean, here's a couple from only the last few days: Louis Vuitton - https://ca.louisvuitton.com/eng-ca/stories/louis-vuitton-via Nike X EA Sports - https://twitter.com/dotSWOOSH/status/1677088133559508993 Mercedes Benz - https://twitter.com/MercedesBenzNXT/status/16635857206892052...

I don't doubt that most of the "NFT Drop" things are worthless money grabs, but my point is that the tech is still darling to tons of corpo execs, and so lots of people in the industry are still getting $$$ to make cool things :)


What is Louis Vuitton's site doing? All CPUs at 100% load to get 2 FPS animation of a textured cube? It looks like a Flash site from the 1990s, but with far worse performance.


Trial projects.


I joined a blockchain startup late 2021, moments before the crypto winter broke out.

I left 15 months later, but they're still going, just did an alpha release.

Because of my work being present on GitHub, I've been contacted by recruiters over email for similar positions several times this year.

I went to two conferences on blockchain and zero-knowledge. Everyone at these conferences seem like serious people working for funded companies, and everyone seems to be okay with there being no customers. It's a little surreal.

It seems that some of the early profit makers in the space are still sinking money into the hope of a future "after the crypto winter" bubble. There is a lot of zero-knowledge hype. Some of the startups that claim to use zero-knowledge cryptography simply aren't.


Based on my LinkedIn inbox… plenty of them still around & still hiring.

They just seem to be less about what was original called “web3” and more about applying crypto to… stuff.


I also still get seemingly pointless pitches about crypto … businesses(?) looking for engineers. Not one have I read and thought hey that sounds like a real business plan


Yeah, most of them are—at best—an interesting feature looking for a business


Web3 makes more sense than cryptocurrency IMO. Instead of applying all the computing power to printing and gatekeeping imaginary money, why not apply it to well, computing? Web hosting and running calculations is something of actual tangible value.


I would imagine most people working for Web3 startups were doing it for the money, and are probably not too surprised that the gravy train hit the buffers eventually.


I was doing a lot of contract work a little over a year ago for about 3-4 different web3 clients. I’m back to working FT at a regular startup. The money isn’t there any more and that’s all that anyone who actually saw what was happening really cared about.


It's a cycle, I was in the last wave in 2018. At some point most developers would realise that chain-scaling issues would prevent profit. They either left to go work for FAANG, or transitioned to working directly on scaling. There are a multitude of grants and other funding from the Ethereum foundation, Gitcoin, DAOs, etc that facilitate a comfortable life for anyone who wants to work on these things.

Nearly all Web3 startups are fully remote, so once the CTO leaves they just sort of collapse of their own accord. CEO is usually a loony.


We're doing fine.

We're in long-term, so scams and hype are something that we try to ignore as we build for practical use cases. But yes, the days of free venture money are over. These are now flowing in the direction of a different two-letter hype.

Source: I am a founder of a web3 startup that has survived a couple of crypto winters.


Some of these scamsters have moved to Web5 https://www.tbd.website/blog/tbd-web5-tech-preview


I’d say that 95% or more is just scam. People like Sam Bankman Fried. They are all gone or dying now more or less. The strong and real businesses with profits and sustainable revenue models are still up and running. There are quite a few examples. Specifically in web3 (within the blockchain and crypto sphere): https://www.coinbase.com/web3/dapps/swappin-gifts is a real world utility web3 dapp. Maybe the first and only one today..


Swappin.gifts is a dApp that allows users to purchase from a selection of thousands of different gift cards using thousands of different crypto currencies over several blockchains. Built on web3, swappin.gifts acts as a new and transformative off-ramp solution.

So it's a website that exchanges crypto tokens to overpriced gift cards. Why does this need web3? This could be 3 PHP files on a VPS.


Why anything at all needs to be web3. You have to ask yourself this question. It’s quite a deep rabbit hole - you have to understand some economics and governance in macro in order to see the value. The other way to see the value is if you are unbanked - like a large percentage of the world. If you are in this group, then blockchain technology (and crypto) is a life changer.


> It’s quite a deep rabbit hole - you have to understand some economics and governance in macro in order to see the value. T

Demagoguery instead of a direct answer about the actual business

> The other way to see the value is if you are unbanked - like a large percentage of the world

Crypto does literally nothing for the unbanked. Zilch. Nada. Zero. It's a playgorund for rich westerners.


He asked: “why this needs to be web3?”

Most people, even these that work in the sphere don’t know the answer to this question.

If you want to know about the business then check their website.

If you want to discuss crypto at large then as I said in my comment, you gotta understand the foundations and not just have an opinion.


Also, that looks like money laundering...


Op was about web3 startups, not about why web3 is relevant. I actually use swappin.gifts occasionally to buy Apple, Amazon. It’s very convenient for people that already hold crypto and want to use it.


> Overall ranking#495

> Ranking in Swap#125

> No ratings

Sounds like a random dapp to push... Are you the author or related to swapping-gifts?


I’m from this sphere. I checked a lot of dapps and when dapps even work, most of the time it’s either gambling, or something completely of no value in the real world. Therefore my comment.


The joke of it is that Web3 claims to be the answer to decentralization... but in every case of Web3 there is some reliance upon a block chain. If the block chain is limited to only the client and not on any server, therefore unnecessary and worthless, it would then be not centralized.

I get the impression nobody takes the time to actually parse the word decentralization. It only means not centralized, but that applies to absolutely everything in the scheme. Somehow people get this confused and for whatever reason believe if anything is not fully centralized then all of it must be fully decentralized.

In order to achieve actual decentralization with online or web technologies consider this incomplete checklist:

* No DNS

* No identity manager

* No shared server. A server is fine so long as you are the only client that can access it as a server or other clients are currently known and prior authenticated (a peer).

* No root certificate or certificate authority or CRL.

* No blockchain

* No shared database or other shared services

Otherwise claims of decentralization are probably just scams to pilfer your data, steal your money, or mine crypto remotely using your hardware. Decentralization is still possible as I have proven it in a personal application of mine, but its very complicated and extremely different from how the web works.


I have a solution to this in https://github.com/anacrolix/btlink. Addressing your specific points:

> * No DNS

It extends DNS by mapping naming over DHT, and content addressed data to a corner of the existing DNS system (.btlink is the working example given).

> * No identity manager

Yep

> * No shared server. A server is fine so long as you are the only client that can access it as a server or other clients are currently known and prior authenticated (a peer).

It allows for a level of shared server to suit the user, to ease use. You can run your own "server" (the eventual default), or use a public or shared one to get started. Server here refers to a process that extends DNS and serves resources without having to modify existing software.

> * No root certificate or certificate authority or CRL.

It makes no attempt to modify the existing system in use on the internet. A certificate is required to serve .btlink traffic over https.

> * No blockchain

Yep

> * No shared database or other shared services

Yep (I assume a DHT doesn't count here).


My goal is to accomplish Zero Trust Conformance[1]. You cannot do that using web technologies in their current form unless you internalize the centralized aspects of those technologies or validate them under contract.

[1] https://learn.microsoft.com/en-us/security/zero-trust/zero-t...

That is the Microsoft definition. The US Army is forming its own definition because aside from reliance upon cloud vendors as service providers the military owns its own infrastructure and services (OSI layers 1-7). Nobody else can claim that, so the US Army can take Zero Trust to a different level than others are afforded.

My goal is to conform to the more demanding military definition for OSI layers 5, 6, and 7. The business case is two-fold: privacy as a mandate and mitigating disruption via logic redundancy. I have an application that is 98% of the way there. I only accomplished this by either internalizing some technologies and abandoning those I could not.

As for certificates that is something you can internalize. You can create your own self-signed certificate and corresponding certificate chain necessary to make your OS and browser happy. If you don't need certificates for their actual purpose, trust authority, then all you need them for is facilitating TLS traffic. I am exchanging trust for authentication via identity and key exchange.


Ok, well the certificate model I described for btlink I would considered "internalized" under your definitions.


> * No DNS

DHT with either bootstrapping or literally just rip through every single IPv4 addr and see if it's a peer lol.

> * No identity manager

How do you know what your friends' phone numbers or emails are? Ask them. No centralized DB needed.

> * No shared server. A server is fine so long as you are the only client that can access it as a server or other clients are currently known and prior authenticated (a peer).

Ethereum (and many of its sibling/child evm chains) clients can run on a Raspberry Pi. It's really not that hard to run a node, and could be packaged into common usage easily.

> * No root certificate or certificate authority or CRL.

Yeah, this is the big one for sure that needs to be solved :)

> * No blockchain

What is wrong with a public distributed ledger?

> * No shared database or other shared services

Not sure what this applies to.


> It's really not that hard to run a node, and could be packaged into common usage easily.

In my experience, it's not hard to run a node, but it can be hard to get it to sync with the Ethereum network, especially using older hardware. You need a lot of fast SSD disk space, and the blockchain keeps growing.

Also, nowadays you probably also want Polygon and maybe some other L2 nodes, since Ethereum's scaling strategy is based on layer 2 networks. And they need even more disk space..


> What is wrong with a public distributed ledger?

It is a highly centralized artifact, which is maximally self-defeating if the goal is not centralizing. When you actually achieve decentralization nobody else is storing or accessing your activity or hoarding your data. There are no public ledgers or anything to crawl or anything to screen scrape, because its on your hardware and you determine who can access it.


Vitalik touched on this topic some years ago.

"Blockchains are politically decentralized (no one controls them) and architecturally decentralized (no infrastructural central point of failure) but they are logically centralized (there is one commonly agreed state and the system behaves like a single computer)"

https://medium.com/@VitalikButerin/the-meaning-of-decentrali...


* No bits

* No wires

* Direct brain synchronisation only


Pivoted to LLM I expect.


Web3 is related to NFTs? Because I feel that NFTs where the most mainstream hype that now is buried below the AI hype. Remember those monkeys?


At the core of Web3 projects is the idea that one can own their identity via cryptographic proof.

Over time, this concept will find its way into many projects. Especialy into open source projects which try to make the web a better place.

For example, as soon as browsers support a DNS based on cryptographic proofs like ENS, other technologies which have URL based identities (like ActivityPub) will automatically support cryptographic identities. Which would bring it to projects like Mastodon and Bluesky automatically.


Still around. Render looks good for the VFX industry as does Helium for broadening 5G reception. Homebase does tokenised real estate that trades immediately for close to nothing since it’s crypto. Lots of payments international payments companies obviously.

Hacker News just generally doesn’t keep up to date because they think crypto is scams. That’s like thinking generative AI is deepfakes.


Be aware that Helium is for sure a massive scam. [0]

[0] https://www.forbes.com/sites/sarahemerson/2022/09/23/helium-...


The perception isn't there for no reason. There is a lot of fraud, but also a lot of product with no market in the space.


AI (remember the period a few years back when everything was briefly a chatbot startup? Microsoft Tay era) -> blockchain -> web3 (blockchain renamed) -> metaverse -> AI (LLM era).

It's the circle of hype... There'll be something else along shortly.

(Some of the surviving web3 startups literally seem to have rebranded into LLM startups.)


Anyone who has seen what happened to Multichain lately?


Web3 is getting rid of stupid scams to keep its foundations solid.

In the long run it will benefit everyone.

However we're probably entering the bull season and we'll see a new wave of web3 scam startups soon nevertheless.


still bullish on crypto, space seems dead but serious folks are still building


There are new ones being made every day. Web 3 is just getting started.


> The term "Web3" was coined in 2014 by Ethereum co-founder Gavin Wood

We’re eagerly waiting!


I've heard it's just getting started for years. When does it actually achieve something of practical importance?


Immediately after cold fusion :-D


10 years later is absolutely not "just getting started". We built Web 1 in 5 year and web 2 in 5 years, complete, with uncountable real businesses generating actual profits and user value. Web 3 has had far more time and far more money and people invested than Web 1 and 2 combines and has yet to demonstrate any real business or user value. What a joke. What a scam.




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

Search: