AI Network
Published in

AI Network

AI Network Monthly Update: May 2021

Engineering

Blockchain Core

AI Network blockchain adopted measures to prevent abusing our network and to enrich our ecosystem. These measures include gas fees for the transaction bandwidth (i.e. number of set operations) and state tree size budgeting.

Bandwidth gas fees are charged differently for different types of transactions. For app transactions that modify values, rules, functions, or owners at `/apps/*`, no gas cost is charged directly as of now. However, the app’s total staking value is taken into account when forging a block. More specifically, an app is allocated bandwidth proportional to its stake, so when App A has staked 20 AIN and App B has staked 80 AIN, App B can use 4 times as much bandwidth as App A. For service transactions, or non-app transactions such as /transfer, /staking, /escrow, etc., gas costs are collected through _collectFee native function. Gas cost is calculated as gas fee * gas price, and a user can specify the desired gas price in microain (= 10^-6 AIN) by setting the `gas_price` field in the transaction’s body.

State budgeting means we now have a maximum allowed tree size for our entire state database as well as for each app. We’re now experimenting to find the optimal sizes that will keep the state tree maintainable without sacrificing usability.

Partnership

POZAlabs X AI Network

AI Network announced a partnership with POZAlabs, an AI music startup based in Korea. A collaboration between AI Network and POZAlabs has resulted in the first AI music subscription service on the AI Network blockchain. Let’s get on board the path of AI-generated content with NFT! The collection of AI-composed songs is now on OpenSea.

For more information, please read the full blog post below: https://medium.com/ai-network/announcement-ai-network-issues-artificial-intelligence-composition-music-nft-by-pozalabs-6dfb9e3af15b

Marketing

Event: Ainize’s Teachable NLP Challenge

Along with Teachable NLP’s launch, the challenge also began. During the Teachable NLP Challenge, from May 5 to May 23, 2021, anyone could train their AI(NLP model) without coding for free. Our challenge garnered a lot of interest from across the globe, and we are proud to share submissions and the winners’ models!

How did the challenge go? The following three NLP models were created by challenge winners.

🏆 Grimm’s Fairy Tale for my child

🏆 Make Top10 Model

🏆 Someone’s Comfort

See more submissions in the Showcase category of the community forum.

Updates from AI Network SNS Channels

YouTube: The open-source community is full of AI lovers and their thoughts and enthusiasm towards it are quite impressive. The AI Network team has been publishing their stories every two weeks on our YouTube channel since last month. We call them “AI Stars”. Each star will share his or her experiences with Ainize and open-source AI.

Write a novel with Kant AI (Eunji Lee, AI Star)
Chatting with AI, OpenChat (Hyunwoong Ko, AI Star)

Medium Blog: For the purpose of sharing insights and enhancing knowledge about open-source AI models with everyone, we’ve published a series of articles titled “Everyone’s AI” on our Medium blog. Explore various articles from May!

Clubhouse: In the clubhouse, the AI Network Team hosts a weekly/biweekly discussion session to discuss current AI issues. Andrew Ng’s data-centric AI seminar was on the agenda for May 15; Timnit Gebru’s AI Ethics and Datasheets for datasets of Google and Microsoft were on the agenda for May 22. The AI Network Club is open to the public, so keep an eye out for us!

That’s all for what’s happening in the AI Network. See you next month!

If you want to know more about us,

--

--

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store