/biz/ - Business & Finance

>>6512292
>10x

>>6492781
>They are building an oracle for SWIFT
>It has nothing to do with Chainlink

Pic related was the main slide in Sergey's Sibos presentation.
Count how many times it says "Chainlink" in there.
Spoiler: six times.

>>6415659
100% LINK

>>6330248
You need a common framework to make even remotely complex transactions like pic related work.
Imagine if every party to this transaction used their own in-house oracles; it would simply not work.

>>6232159
>hurrr banks will build their own oracles
Good luck getting pic related to work without a common framework.

>>6114293
Swift is already a centralized third party used by banks. Adding another on top of that would be pretty dumb.
Plus, it's a lot cheaper if you have a free market of nodes vs a centralized single party who holds the monopoly.

>>6114645
Pic related was the main slide at the Sibos demo.
Count how many times you read "Chainlink" on there.
Spoiler: six times.

>>5807048
The Swift smart oracle isn't a "private oracle system" you mong.
See pic related, it was the main slide in Sergey's Sibos presentation.
Count how many times it says "Chainlink".
Swift's smart oracle is simply part of the Chainlink network.

>>5757444
Also, Swift's use case for Chainlink is pretty much pic related.

You have a bunch of different banks, an external party like S&P, and Swift all exchanging information through the common CL framework.
The way this would've worked without PSD2 is all of those parties would choose to open up their APIs and data feeds to the CL network.

With PSD2, the process would be the exact same, except there would be a lot more completely independent nodes to choose from for nearly every part of this flowchart.
This only enhances the decentralization, and thus the security, trustlessness, and overall health of the network.

>>5218270
Oracles are pretty common things. Companies set up their own all the time.
The benefit to Chainlink is that it provides a common framework where everyone uses the same oracle solution, which means that every party to more complex transactions knows what they're getting.

See pic, which is the use case example developed for Swift, as shown at Sibos.
If every party to this smart security (banks, S&P, Swift) used their own proprietary oracles, none of the other parties would have any reason to trust the other parties' oracles.
Plus, obviously the decentralized nature of the Chainlink network also adds a lot more security (through redundancy etc.).

In other words: if this isn't about Chainlink, it's likely a placeholder oracle solution, and they may very well decide to use the much superior Chainlink in the near future.

>>5178595
>>5178898
Also, I think the Sibos use case was such an internal process, or more precisely a collection of such internal processes (and a few non-internal ones, like those interest rates perhaps).
But as you can see, even if all of the individual processes that make up pic related were 100% internal; you can see that there are a number of different parties involved. So if everyone were using their own internal oracles, nobody would have much reason to trust the other parties' oracles.

>link shall rise as btc falls

>>4727574
What can you tell me about pic related?

>Chainlink nodes need open APIs
>EU is about to force banks to make their info available through open APIs in 2018

This is like forcing farmers to grow dandelions after Sergey invented a car that runs on dandelion juice.
Get ready, lads.

Source: https://www.cnbc.com/2017/07/11/incoming-eu-banking-regulations-are-a-headache-for-everyone-rbs-fintech-advisor-says.html

>>4665263
>they are years from a working product
lol ok buddy

>"the smart contract market doesn't even exist yet"
>"AXA is doing smart contracts".
Lol you dumbass.
What's worse is you typed this back to back.

>it worked perfectly fine with regular oracles
Good luck doing pic related when everyone uses their own proprietary oracles.

>>4576237
>>4576276
You have poor understanding of what Chainlink actually does.

See pic. Variations of it made up half of Sergey's Sibos presentation.

You not only have various steps with a variety of different contributors (banks, S&P, Swift, ...), but you also have lists of multiple contributors (that list of banks on the left supplying the same kind of info).

If every party to this transaction used their own oracle solution, nothing would get done because the other parties would have no reason to trust any of those proprietary oracles implemented by those other parties.

If every party to this transaction used a single trusted centralized oracle, you'd have a single point of failure and a single target for tampering and attack.
You'd also have a very powerful central unit that would have all of the parties by the balls when it comes to pricing etc.

Decentralization is the only way.

>>4535948
>>4536329
>Sergey doesnt mention link
This fucking meme lmao.
The product is smart contracts, which is naturally where the focus lies.
The Chainlink network and the Link token are simply tools for achieving that goal.

And variations of pic related made up half his Sibos presentation. Notice anything?

Retard.

$LINK

>>4463109
>For MOST smart contracts you are dealing with oraclized data from a singular data source.
Source on this "most" thing.

And the point is that decentralized oracles can be used for complex transactions like pic related, and also for simple transactions.

Centralized oracles would only work for simple transactions, not complex transactions like pic related.

It would be impossible to get all the parties in pic related to use an external centralized oracle like Oraclize; those parties would much rather just set up their own oracles in such a case.

>>4411138
>they have no use at all

>>4270184
>I would like to see someone debate anything I've said with proof
Gladly.

Look at pic related.
This slide (and variations of it) made up half of Sergey's presentation at SIBOS.

Notice how you have a series of banks from which information is drawn, as well as Standard & Poor's.

There is only node directly communicating with SWIFT in this example.

For this transaction (smart security) to have any kind of legitimacy, the oracle computations need to be as trustless as possible, for everyone involved (those banks, S&P, Swift, Security buyers, ...). That's one reason for the decentralized network (and thus the token).

Another reason is this: with a decentralized network with built-in trust features, SWIFT would not have to set up a new node every time a new external party is added.
And SWIFT also wouldn't have to keep nodes online for all that time they don't actually need it or its data.

Make sense?

If you can poke holes in this, I'm unironically all ears.

>>4238140
Good reply:

>I could write Oracles while taking a dump.
That's the whole point, big boi.
Companies and banks and shit out there are currently writing their own oracles for their specific shit, like AXA Fizzy.

Now look at pic related (from Sibos).
And now imagine every single party to this transaction using their own in-house oracles.

The whole set-up would be doomed to fail, since nothing is connected, and the trust and confidentiality levels are wildly different from one oracle to the next.

Chainlink connects everything. It's a decentralized network; the "internet of oracles".
And even for very simple oracle computations, Chainlink will be a vastly more robust solution than any in-house made oracle. Not to mention cheaper.

>>4231034
It would be more cost effective and more secure if they did

>pic related

>>4228312
>I could write Oracles while taking a dump.
That's the whole point, big boi.
Companies and banks and shit out there are currently writing their own oracles for their specific shit, like AXA Fizzy.

Now look at pic related (from Sibos).
And now imagine every single party to this transaction using their own in-house oracles.

The whole set-up would be doomed to fail, since nothing is connected, and the trust and confidentiality levels are wildly different from one oracle to the next.

Chainlink connects everything. It's a decentralized network; the "internet of oracles".
And even for very simple oracle computations, Chainlink will be a vastly more robust solution than any in-house made oracle. Not to mention cheaper.

>>4193672
>WHY THE FUCK would a company need decentralized oracles
The whole point of using blockchains is their trustlessness, which is itself a direct product of decentralization.
If you run external data through centralized oracles in order to tie them to the blockchain, it defeats the purpose of the blockchain.

This is especially applicable to cases like pic related, with multiple parties and multiple similar sources.
However, Chainlink is also ideal for very simple oracle computations, with a single source for instance, because the system is so robust (off-chain security with SGX, node reputation, etc.).