On August 1st, a subreddit appeared where the astronomer from the Space Telescope Science Institute and two blockchain teams, which are using blockchain to help categorize millions of deep space photos from the Hubble Space Telescope, started an interesting ask-us-anything session.
Josh Peek, associate astronomer, Data Science Mission Office Project Scientist at STScI, and MAST Principle Investigator, under the nickname “carbon_is_metal”, answered a variety of questions.
Astronomy Normally Demands a Lot of Expensive CPU Working Hours
In one of his first answers, Peek explained that they basically “cut the sky up into small images that fit the pre-trained network: 224×224 pixels,” before stated that there are around 100 TB of Hubble images, and due to the way the algorithms work, they need to send them in a lot of different ways; zoomed, flipped, etc, meaning that they could be looking at tens of millions of CPU hours and up.
He then stated that so much CPU hours would be incredibly expensive on the commercial cloud, but with their partners, Hadron and Aikon, which are devising blockchain solutions, the prices get substantially reduced.
Aikon and Hadron to Improve the Performance
Aikon helps decentralized developers to monetize their hard work effectively, while Hadron improves the accuracy of AI models by partnering up with the community and allowing them to run the computations on their idle computers and devices, which enhances the AI quality and reduced compute costs.
“That’s what makes what Hadron is doing so cool – they are built for this kind of AI implementation. In fact, they have a few neural net folks on staff, so they actually helped us tune the network for our particular purposes,” said Peek.
Mr. Peek also deduced how astronomers can also use regular laptops to process all the needed information, but with such a consumption of time that it couldn’t be considered efficient. Therefore, the team has turned to the blockchain technology to speed the process up.
Two Blockchains in One System
The Founder of Aicon, Marc Blinder described: “In this case, Blockchain allows large groups of computers to work together efficiently and quickly exchange value. That enables new business models that lower costs,” and differentiated two blockchains that are used on the project.
“Aikon uses blockchain to manage access rights and payment for APIs, while Hadron uses blockchain to divide and organize large AI tasks and pay for the computing power of individual machines,” he deduced.
Blockchain is a Supercomputer
This three-way partnership is using the blockchain technology as a form of a supercomputer which has the capacity to store and transfer a large amount of data more efficiently than any server.
Which reminds us of the Conversations With Tyler podcast with Vitalik Buterin, which was recently covered by theblockchainland.com, where the Founder of Ethereum (ETH) compared blockchain to the average computer by saying:
“It has a hard drive on which it stores what all the accounts are, and what the code and the memory of all these smart contracts are. Blockchain accepts incoming instructions , which are signed and sent by a bunch of different users , and processes them according to a set of rules.
On a blockchain, you can ultimately build anything that you can build on top of a computer,” Buterin explained and concluded that “From a computer science theoretical point of view, in terms of what it provides, you can think about the blockchain as being a computer.”