I think we can work out the income you could make from a node based on the number of monetised API requests that are...

I think we can work out the income you could make from a node based on the number of monetised API requests that are being made nowadays by traditional businesses and the amount they charge.
>nordicapis.com/the-ultimate-guide-to-pricing-your-api/
Assume that smart contracts requiring external data become mainstream and the crypto API economy reaches the same size as the regular API economy now.
Let’s take the “hobbyist” from nordicapis.com as equivalent to a NEET running a chainlink node. That’s 688,991 calls per month. Per year it’s 8,267,892 calls. Say you charge $0.01 per call, which nordicapi reports, and is the minimum Oraclize charges. That’s $82,678.92 a year. IBM Watson charges $0.0025 per call which would be $20,669.73 per year.

Docusign says for their API that "You may not exceed 1,000 API requests per account per hour". So 1000 per hour is 24,000 a day, times 1 cent per job is $87,600 per year per node operator. With 19,000 node operators (Sergey's number) serving 1000 API requests per hour that's a total revenue of all nodes of $1,664,400,000 per year. That’s only 0.08% of $2.2 trillion (estimated total value of API economy in 2018).

Attached: 9E761ADB-E4CC-4B62-9DC2-85166B805C20.jpg (1800x2048, 596K)

Other urls found in this thread:

blogs.wsj.com/cio/2018/03/21/the-morning-download-salesforces-bets-6-5-billion-on-the-api-economy/
twitter.com/NSFWRedditVideo

Salesforce just bought, for $6.5 billion, Mulesoft Inc., a company that's about making APIs externally available:
>blogs.wsj.com/cio/2018/03/21/the-morning-download-salesforces-bets-6-5-billion-on-the-api-economy/
This is the most expensive acquisition Salesforce has ever made. That's because they know how important APIs will be, and ChainLink is poised to connect this API economy to this emerging Smart Contract economy. Salesforce are connected to ChainLink via the dev Dimitri Roche, and Sergey tweeted the Salesforce CEO.

Yes, the API economy is huge and it's growing. How huge? Well, the research consulting company Ovum says that the size of the API economy is going to be 2.2 trillion dollars during 2018. There were ~15,000 public APIs available in 2016, growing by 40 new ones per day during 2015. In 2016, Netflix alone received 5 BILLION API requests PER DAY.
>An API request is equivalent to a "job" that an oracle would do.
>For comparison, Oraclize claims to do "thousands" of jobs per day. That shows just how much room for growth this API/Smart Contract economy has.

If ChainLink can capture just 3% of the $2.2 TRILLION global API economy, that gives it a market cap of $66,000,000,000 - 66 BILLION. Consider that Sergey said there are 19,000 people interested in running a node: that's over 100% of the available public APIs in 2016, so 3% of the market is very conservative. Consider that Starbucks holds 40% of the market share of US coffee shops. We’re not even considering what % of the smart contract economy ChainLink can capture, for which I can't find much on predictions for its value or its current value.

Price per coin multiplied by number of coins determines market cap, so marketcap divided by number of coins gives us coin price. So if LINK captures 3% of the API economy, that gives us a price of $188.

Attached: 20B57EA0-CBCA-4C6C-A0B1-1ED3682D9D27.jpg (1000x1080, 233K)

Attached: 1536959237631.webm (512x384, 2.69M)

All that’s very interesting. But this is just about the API economy. Sure, we can make estimates on link price and node income based off that. But what about the smart contract economy? How big will that be? We don’t know. But what we suspect is that smart contracts will slot right into financial markets. That’s where pic related comes in.


This was a pasta/infodump of price speculations based off clearly stated assumptions (like api economy size, number of requests per day, etc). If you have any more estimates like this then please share.

Attached: B3E2DADB-E54E-43C2-A775-C73E6C9F9943.png (756x1012, 135K)

What about high value contracts?

See

the thing with link tokenomics is the value of contracts.
i still can't wrap my head how can link even function without a huge spike in its token price

lets imagine the network is running and some gigantic contracts start appearing in the network, with big penalty fees. either the price is high enough or the link supply could be immediately concentrated on just a bunch of nodes.

Holy shit i am selling my house, a kidney and a testicle. Per pee poo poo

The combination of both these things is what makes link so well placed right now. We got APIs - sheer volume. We got financial markets - sheer size. There are a lot of link tokens but are there enough for everyone in the financial markets and api industries that would benefit from adopting smart contracts? (Hint: no)

A single link is divisible to 18 decimal places - there are plenty of LINK to go around!

In my opinion LINK could also get a piece of the cake in the supply chain sector and government (identities? Notary stuff?) also. What i mean is that it could go the other way around and create new doors for the API economy.

Can it also operates within API services? One API requesting something from another one that needs to be secure? Does it makes sense?

Nice OP I’ll take another 10k

Attached: 1537134278426.webm (960x720, 1.71M)

>with big penalty fees
you're assuming like it's a given.
What if the contract creators won't give a damn about penalties because they will manually choose the best nodes from the top of the lists that reputation providers will display?
I coul totally see a scenario in which contract creators would rather use SWIFT/Dcusign/some other established entity's nodes with no collateral because they already trust them instead of choosing some no name neet's node with collateral

Hahaha this was really good. Thx fren.

Why would you choose nodes with no collateral? Are SWIFT/DocuSign/other established entity's foolproof? Do you trust them 100%?

of course you trust them.
companies DO trust eachothers, they can be worried about some 3rd party interfering in the process but they sure aren't afraid that the other side will simply fuck them over deliberately. The entire western economy is based on trust and reputation.

I'm realitively bullish on link but if you guys think that we're going to 1keoy because everyone will be running nodes with high collateral making bank I have bad news for you.

Most companies don't even want to use ethereum, they want to use permissioned blockchains - so much for "decentralization"
And link won't be a binary solution that forces you to choose between total centralization and maxmimum decentralization, it will allow for whatever level of decentralization you feel is required. So far it doesn't look like the traditional indsutries are in love with the decentralization meme

And I know what you're gonna say
>b-but what's the point of smart contracts if it's not fully decentralized heh
then explain to me why do companies even consider hyperledger instead of eth? Why will chainlink allow manual node selection instead of choosing it randomly for maximum decentralization?

Have you ever wondered why the Link Pool guys decided to go through this entire hussle of developing it instead of just buying a fat stack of the tokens and patiently waiting for retirement?
Probably because the tokens aren't going to the moon like we would all like and they know it. There will be money to be made running nodes, but it's gonna be the same as todays mining, you mine and you sell it for fiat to secure profits, which won't put any upward pressure on token's price
If link went to the moon and every node operator got rich, they would all quit that shit the next day.

So you would trust a company 100% not to fuck you over in a transaction? Just want to be clear here

Node fails because fuck reasons, well tough shit jim, you just lost money, even so trust me like you always do

> People will do what I say they will because of reasons

Attached: 1520850849598.jpg (469x584, 494K)

What about if the price of link falls during the contract? will the node have to stake more?

According to Thomas, that is not the intention. However, nodes probably won't have to put up collateral for significant periods of time, therefore lessening the chance of major price changes

Contract will take literally seconds to execute. Effect will be Negligible

There will be so many smart contract creators that there will be plenty of contracts to go around for NEETs as well as banks and whoever. If banks don’t want to choose random nodes then fine. Someone from the myriad industries link will transform will want random nodes.

Can Iget an explanation ofhow to run a link node?

thanks user

I sure would trust that Swift won't fuck me over deliberately which would be transparent to everyone. Sure there is a chance that one of their nodes gets compromised by sme outside bad actor/downtime/whatever but if they themselves run multiple nodes in an already "decentralized way" I don't see the problem. Add a few more nodes from the #2 and #3 from mr Shekelsberg's reputation service and you're good to go.

You're angry because it doesn't scream 1kEOY!!!!!!!!!!!

sure it's possible but it's not a sure thing like all those people claim, daydreaming about some calculations that link market cap must = this or that because of XYZ industry will "require" this or that much collateral

So you've gone from having 100% trust to now saying "Add a few more nodes". Kinda weird thing to do if you have complete trust isn't it?
Also, if you only use one node how do you know they fucked you over? What are you going to compare the result against if you have no other nodes providing data?

Has anyone setup a link node? I can't wait for the threads called passive income general

screencap this

Attached: 1532274182059.jpg (720x960, 127K)

>So you've gone from having 100% trust to now saying "Add a few more nodes"
a few nodes of each one of more nodes of one particular entity, just to prevent potential downtime.

>Also, if you only use one node how do you know they fucked you over?
One entity can run multiple nodes

£342,857 LINK eoy

Attached: 1524889370287.png (658x627, 342K)

Dollars not pounds you euro fag

But if it’s the same entity then all the nodes would return the same value? Why would nodes from the same entity return different values? They’re basically fucking themselves over doing that.

Mining used to be passive income general a year ago. Look how that turned out. I believe in LINK long term, but if people start speculating and buying it up to run nodes, I hope you anons will be smart enough to sell a portion of your holdings.

can you explain further? what happened to the passive income. i wonder how many links you need to run a node like... the more links you have the more calls you can take? hmmm speculative

>But if it’s the same entity then all the nodes would return the same value? Why would nodes from the same entity return different values?
youve completely misunderstood what an oracle is.
the oracle ensures that a random or deliberate error in data on one node can be picked up by matching with the result of other nodes in the network.
oracles hve no say in whether the data is "true", just that what was sent from one end is the same that which is inputted onto the blockchain

Mining incentive decreases with an increase of miners. Price per node does not decrease with the amount of stakers.

Read the thread first please, you’ve totally missed the point. We’re not discussing oracles in isolation here, we’re discussing why you would or would not decentralise your oracle providers.

you're assuming that all SWIFT's (in this example) nodes would get hacked and data compromised. Big node operators will surely decentralize/run their own nodes in a way that prevents that. But even if it happened the nodes wouldn't get penalized even if they had collateral because decentralized oracles don't know what's true and what's not it's about the majority vote basicly

And if we assume that the data source gave wrong information then it doesn't matter what nodes you used since they all will use the same sources.

Besides that It's mostly about the downtime, not malicious behavior

No, I’m not assuming that all of SWIFT’s nodes get hacked. You are assuming that SWIFT would never fuck you over, that you trust them 100%. But you’ve already back tracked by going from using one node to multiple nodes. Now if SWIFT did want to fuck you over, why would they only have one node send bad data? That would just shoot themselves in the foot. They would have all nodes send bad data.

Have you ever read an ISDA or a bond document or any form of agreement between companies? They always contain clauses around non-compliance/non-performance/negligence because companies DO NOT TRUST EACH OTHER.

no, youre mistaken b/c of this comment you made:
>But if it’s the same entity then all the nodes would return the same value? Why would nodes from the same entity return different values? They’re basically fucking themselves over doing that.
the "entity" could send corrupted data b/c of an error in message processing. it doesnt have to be deliberate. but all the same, an oracle isnt there to confirm that data sent is "true" - it could be completely false - but the oracle will confirm that the data received was the same as the data sent. that's a big difference.

And I’m 100% correct because of the context of the comment. If you actually read the thread, you will see that I’m arguing for the decentralisation of oracle providers when entering into a smartcontract. The reason for this is if the oracle wanted to fuck you over, it doesn’t matter if you had 1 or 100 nodes from that provider since they would make all of the nodes return bad data. Read the thread honestly, understand the context of the conversation

>Now if SWIFT did want to fuck you over, why would they only have one node send bad data? That would just shoot themselves in the foot. They would have all nodes send bad data.
and it would be visible for the entire world to see since you will see that they deliberately changed the data that came out of a verifiable source.
I would rather believe that a meteor hits us all tomorrow than an established company with years of history and billions in revenue deliberately cheats

>But you’ve already back tracked by going from using one node to multiple nodes
I didn't backtrack out of anything, let's say you trust swift the most, you use their nodes, or you have 3 companies you trust equally then you use a certain amount of nodes from each company. Pretty simple

There are many alternative narratives user many of which involve risk. How can you be risk averse in today's world?

Attached: 1535046412319.jpg (1024x788, 139K)

How would it be visible for the world to see? Lol have you ever traded derivatives or anything before? If you ask 10 banks for the 1pm GBPUSD rate you are gonna get 10 different answers! Lol and you think someone looking at what SWIFT provides is going to be able to clearly identify that they didn’t shave a few pips off the rate. If you had experience in the market you would understand that it may be incredibly difficult to prove that they fucked you over.

You know about the LIBOR scandal right? Bank traders in collusion to rig the LIBOR rate. You know, banks with all that history, colluding together to make themselves more money. But they’re 100% trustworthy in your mind.

ok, fair point.

lol nobody is gonna use nodes of entites that are invovled in the contracts because of an obvious conflict of interest.
I used SWIFT just as an example.
You're acting like interest rates is the only use case here
What I was saying is that you cannot just tamper the data and then point at the API because people will be able to verify that.

You’re still trusting one provider though, and hoping that they don’t collude with your counterparty.
Again, how are you going to verify that they fucked you? You are saying “just look at the API” but that means you’re gonna have to have a subscription to that API. Ever subscribed to Bloomberg for FX rates? Do you know the cost? Do you understand how Bloomberg aggregates their data? Is the data that you need actually officially published, or is it a snapshot at a point in time? You’re trying to make it sound like it’s so easy, but it’s really not. If you have spent any time in a corporate or bank environment trading financial instruments you would understand that things aren’t quite so straight forward

So if I stake 10k linkies how many linkies could i make in rewards per week?

are you telling me that if a node operator tampers the data and then makes excuses that it actually wasn't compromised by him but the original API there will be no way to verify that? it's gonna take a second to expose that node operator to the entire world and destroy his reputation.

At least 777

I’m more wondering how you’re going to know that the node operator has returned bad data. Tell me how you automatically know a FX rate is wrong. If I said GBPUSD is 1.3072 now is that right or wrong? How would you verify it?

>If ChainLink can capture just 3% of the $2.2 TRILLION global API economy, that gives it a market cap of $66,000,000,000 - 66 BILLION.
Fuuuuucking hell how can people be this retarded. "Capturing" 2% of the requests does NOT necessarily mean you get 2% of the money.

Imagine all the houses on your street are for sale, and all-together they are worth 100M. If a realtor sells 50% of those houses, does he make 50M?
FUCK NO

Your profits aren't just market value/your share.

Bumping threas cause full copypasta fud is being postre right now

No one is talking about it in terms of truth/false, the only question that matters here is does the answer that the node gave match the data that it was given by the API? if no, then it's quickly verifiable and everyone knows that the node got either hacked or lied.

You're talking about the data coming from the API providers/ data sources. That's not the node operators' problem.

Also for varying things like interest rates I guess there will be some range of acceptable numbers but it's still unknown until the consensus mechanism is out.

Fine, prove it. Prove to me that the GBPUSD rate from Bloomberg at 12:54:12 was not 1.3072. I want you to verify for me through the API that the rate above is wrong. Pull the historical data for that exact second from Bloomberg, let me know how you get on.

Show me I haven’t tampered with the data coming from the API. Because that’s the data that my node sent your smartcontract, so how are you going to prove me wrong?

this but unironically.
also that is yearly api economy, so if the nodes are making 66B per year then the market cap would be at least 660B, otherwise nodes would be doubling their link every year

Attached: 1537045420429.webm (640x480, 2.84M)

Sorry I don't have a bloomberg subscription. Are you saying it's impossible to go and check or request historical data from their API?

its not only about trust you brainlets, its also about getting away from the paper bullshit that is plaguing us, chainlink is literally the way in becoming a paperfree economy and gaining in time, profit and transparency, everyone who doesn't adopt it will lose customers and profit to the ones who do.

Attached: 1534207387903.jpg (1152x6352, 1.29M)

Attached: 1516244611550.jpg (380x405, 12K)

Ozarks wife

also forgot to add that reputation providers will likely gotta have the access to the APIs that node operators listed on their service use so it wouldn't probably even require checking back later, it would be visible immediately that something was off

For that precise second it is impossible, mainly because FX rates can move multiple times within each second. Bloomberg only “officially” publishes historical rates every half hour period. Anything in between is an estimate based off their 1min graphs. Remember, this is all before we even get into the discussion around whether the rate is a bid/mid/ask rate, and which composite (group of banks) it is.
So I could shave a pip or two off the rate and you wouldn’t be able to prove that i tampered with it. I could even probably do up to 10pips without raising any eyebrows. 10pips on a multi-million dollar FX derivative is tens of thousands of dollars I could fuck you out of. Not a bad result for me when I know it’s virtually impossible for you to prove my rate was wrong.

This. The network also becomes stronger against 51% attacks the more people who run nodes.

Having more link allows you to do tasks which require higher collateral.

All I'm saying is if anyone can go ahead and set up a passive income source it won't last for long. It'll most likely cause a huge spike in LINK price if running a node suddenly becomes the most profitable source of passive income, however as more nodes compete for jobs they'll surely drive down the price per task, to a point where only nodes with large collateral can stay profitable.

Be me 100% straight >White, UNCUT ;), also somewhat of a gymhead >Each time I'm done with my workout sesh I have this weird urge to dominate a cute femboy in bed for a couple hours It-... It is normal right? Am I still straight?

if you choose a node operator with no conflict of interest to the contract we would have to assume that this entity was bribed (and later laundered that money somehow) into tampering the data and then hoping that people would take the their side and trust them instead of the reputation provider in case the latter had the same API access and says the node operator was dishonest, everything for a pretty specific use case at that

I don't think even the honest node operators would want to serve API calls that can't be verified later on with the source itself

data providers aren't going to offer api subscriptions to everyone because if everyone has access to that data then its value is diminished. that's going to be the barrier to entry for every lazy asshole who thinks they can run a node and collect a passive income.

It’s totally up to you if you want to centralise your node operator. If you look my example above though, it’s pretty easy for a node operator to fuck you out of tens of thousands of dollars and virtually impossible for you to prove it. But if you are happy to take that risk vs paying an extra few cents per API call for a couple of other node operators to provide you with the same data, well that’s your choice.

holy shit

Attached: toplaffo.gif (279x220, 3.53M)

This is why aggregation isn’t necessary for the network to launch. I was going to make a thread about it but I’ve been busy, but my hypothesis is that the reason Thomas has said they haven’t started on consensus/aggregation is because they don’t need to do it. The network is only to ensure the data sent/received is correct and not tampered with. Therefore, i think it’s possible chainlink is relying on the 3rd party partners to develop their own aggregation/consensus models for the network to use because a) it isn’t something chainlink should develop, and b) they’re the end user customers of those aggregation models which other companies will eventually use.

This also helps explain why chainlink has been so tight lipped about partnerships and relationships with companies: those companies are developing proprietary aggregation models to use when the network launches and don’t want their competitors to see what they’re working on. This also explains why Thomas said they don’t have any hidden coding repositories; because they don’t. The companies and 3rd parties chainlink is working with (docusign, DA, Microsoft, swift, etc.) are all developing consensus/aggregation models on their own to help build the network from a business perspective and probably have those models in their own, separate coding depositories.

Screencap this. Mainnet will launch before the end of 2018 and it will be revealed that all of these suspected “partners” are others developing their own industry specific aggregation and consensus systems because they all require separate and unique ones. Chainlink’s job is to ensure the data from api providers isn’t tampered with, and the companies’ job is to help aggregate the data for the smartcontract products/services they hope to offer their customers.
t. LTG
Hail Sergey

Attached: 1CD7A5CC-D535-459B-9149-3CD918C25343.png (753x1024, 885K)

Anyway, this discussion has probably run its course so I’m off to bed.

>Mining used to be passive income general a year ago.
It still is.

>This also explains why Thomas said they don’t have any hidden coding repositories; because they don’t. The companies and 3rd parties chainlink is working with (docusign, DA, Microsoft, swift, etc.) are all developing consensus/aggregation models on their own to help build the network from a business perspective and probably have those models in their own, separate coding depositories.
I would actually be disappointed in the team if this is the case.
Thomas, in particular, has been very clear in the words he uses, and states pretty bluntly that aggregation work hasn't started. If your idea is correct then he has been deliberately implying an untruth, which would be uncharacteristic.

Nice one. Hope this is the case.

ok so let's say someone picks nodes out of top 3 operators, 2 out of them would have to be bribed to fuck you over and then you have the shit shitting the fan when there are obvious differences in top 3 node operators responses, or do we assume that they all are corrupt.
However you put it I simply don't see contract creators going for the random node choice or choosing lower ranked nodes just because they offer collateral

Agreed. Although if you take only a 10% cut of that you're still at 6.6 billion which is ~17x from here. Seems realistic to me over 2-3 years.

>they'll surely drive down the price per task
value will be perceived by reputation of a node. It seems that the early adopters who run a node will have a significant advantage over new comers who have nodes with no reputation.

I don’t think that’s the case at all. He’s been asked if they started working on it and he said no. That’s true. AND Thomas wouldn’t know what those third parties are working on. Why involve somebody coding the network with having to deal with external organizations and an entirely new problem set (aggregation).

Additionally, something I wanted to add, the data requirements for smartcontracts for the different industries will vary wildly. A law firm looking to aggregate tamper-free data provided by the chainlink network will have very different aggregation and consensus requirements compared to someone using tamper-free data to run banking smartcontracts. The aggregation and consensus for what “executes” the contract will be wildly different. So why would Smartcontract.com be the ones to develop those models?

The only logical conclusion is that those individual companies looking to employ the chainlink network for tamper-free and self executing smartcontract data will want to develop those aggregation models themselves. Smartcontract.com doesn’t have a team of lawyers to develop an aggregation model to resolve smartcontracts lawyers would use. But do you know who does? Check out one of the many legal firms working with chainlink. They do.

It only makes sense for Sergey to outsource the aggregation development to all the “partners” because it differs wildly from industry to industry and they’re the ones who will be using the network and aggregation models when it launches (soon I believe). It only makes sense those third party companies are building the aggregation as well as reputation systems, or doing so in tight conjunction with Sergey/Steve. Reputation will also depend on the industry and what they’re looking for, so I assume that is being outsourced to the enterprises as well.

They have confirmed rhat reputation will be a choice of third-party contracts.
Would be interesting if aggregation were the same.

crank heilung, get high in the lung, heil serg

Attached: 1536767436578.jpg (684x638, 130K)

stinky linkies unite

the numbers involved in the api economy are so huge that at any stage in your calculations you can just knock down a couple orders of magnitude and you're still looking at pulling a share of, at the least, billions of $'s.

LINK will be a licence to print money. The FOMO immediately preceding the singularity will be astonishing.

that makes sense, (consensus mechanism needs to be use-case specific I guess) and if true would make Chainlink an absolute behemoth in the crypto space making every other shitcoin basicly disappear in shame in terms of acomplishments.
If this all comes true CL will be an octopus reaching everywhere with its tentacles

I believe all this, but my only question is - why are we the only ones who can put the pieces together? No matter how many NDA's are in place, there will always be leaks and always be someone buying based on inside information, but in this case the token price is still thirty something cents.


is it because each component is split that no-one can see the forest for the trees?

I think so. Sergey is smart. Sergey knew that if there was only one aggregation system when the network launches (as developed by chainlink) then it would be wholly incapable of aggregating data for unrelated or tough to configure disparate industries. HOWEVER, Sergey is smart. Sergey knows that if he brings in the industry leaders and has them develop the aggregation models they will use, he’s now created not just a decentralized oracle network, but a decentralized oracle network with ready to go aggregation models on day 1 built by experts in those industries so anyone can start using the network on day 1. Or something like that. Just my hypothesis. I’ll need to make a Big Mac run for more thought provoking sustenance later.

I think spending more time exploring rinkeby could provide some better insight into this, but I wouldn’t know where to begin.

I think many, many people (and enterprises) would like to see the network work first (mainnet with real money) before diving in. For speculators Link is a godsend, but for an enterprise there’s zero risk to waiting for it to go live before investing money in it.

kek

Attached: tom.jpg (407x407, 24K)

Same reason thst rajesh cannot distinguish between his neighbor's floater and Pajeet's dingleberries from up the street.

It's because if you really really analyZe these crypto people who think they are so smart/edgy/high IQ/forward thinking.......... are really not that intelligent in fact. They don't have common sense.

Look at some of these people making crypto YouTube channels.... and reddit people.... you can juSt look at them and you can tell they are dunderheads in many instances.

They latch onto 1 or 2 specific coins and are brainwashed. Unable to move or think differently.

I think we are more savvy in the way a drug dealer who makes a lot of money who is savvy. That drug dealer is doing the same thing as some dipshit cocksucker selling insurance but we see it from a different angle.... even if some of us don't really get it.

pee pee poo poo

Attached: B4C4E427-CA88-4D4C-B744-F94ADC208658.png (689x795, 96K)

To expand....

I've never read a white paper of any coin. Because I'm not smart enough to know if I'm right about what I'm reading.

I still haven't set up any wallet because I know ill either not be able to do it or I'll fuck it up someone.

But it's so obvious to see where the world is headed with what I understand Link is trying to do and making it happen.

The one thing that made me invest in Link more than anything else is universal basic income push that is starting to happen. You see about tests on it and you hear it more and more in the news.

In a vacuum from an ideological perspective.... UBI is shit. But then I see what Link is doing and then it made total sense to me why I'm seeing more talk about UBI the last few years.

It's because Link will make UBI a necessity

In what way? Because LINK will enable a multitude of smart contract use cases and displace jobs?

>If

Yes.... you keep hearing about UBI the last few years. And I'm thinking to myself are these people fucking stupid?

Then I get the general idea of Link and I'm like oooooooooh ok that's why these people are pushing UBI.

A true singularity. Everything comes together. The memes. The stinky linkies. Kek. Skynet. Apocalypse.