HUMAN Protocol powers decentralized marketplaces where you can request and contribute virtual work within the world’s largest labor pool.

Today we have Harjyot Singh as Technology & Crypto Director of HUMAN Protocol joining us to answer our questions for the next hour.

Harjyot Singh, HUMAN protocol Technology & Crypto Director

A seasoned engineering leader, he has worked in a variety of companies and industries over the past decade, with a focus on finance and distributed computing.

He has also previously founded two start-ups, tackling issues such as misinformation and privacy.

Question from the CryptoRank team

#1. Could you please tell us more about yourself and your experiences?

Harjyot Singh: – Sure 🙂 I left India to study artificial intelligence at the University of Edinburgh, and I was immediately interested in startups and the cutting edge technologies like AI & Blockchain. Surprisingly, I was good at it, and was a tech lead by the time I graduated.

I built my first start up – Provenance Labs – to tackle misinformation, which involved semantic labeling modelling, not only to see what is fake, but what the bias is in it.

The back of that led me do my entrepreneurship degree at Oxford University & straight after to Human Protocol 🙂

#2. Can you tell us about HUMAN Protocol? What problems do you solve, both in the space and for the community?

Harjyot Singh: – HUMAN Protocol is really simple. It is a software infrastructure that supports the creation of decentralised markets for all kinds of work, with the current focus on data labeling work. A way to look at is this:

With birth of apps like Uber, Fiver, Upwork etc. people started realising that they could utilise their non career skills to solve problems and earn money – there certainly was a demand for it on the other side. This lead to a new generation of gig economy products.

However building the core technology for such kinds of dual sided marketplaces is quite complicated.

At human that’s we have built – the entire pluggable backend to build these marketplaces entrepreneurs, companies can really focus on the marketing & dynamics of the marketplace they might launch, we handle the tech 🙂

We are also rolling out a marketplace built on top of the core Human Protocol ourselves – ML Data labelling marketplace. But more on that later 🙂

#3. Could you shortly describe the milestones HUMAN protocol has achieved, and about your upcoming plans?

Harjyot Singh: – So the first application that utilised Human Technology was HCaptcha – it powers something insane like ~15% of the internet traffic. It’s our first partner application. We are in the process of launching new tools for more granular ML Datalabelling. Intel’s CVAT & Inception .These allow anything from simple images to complex videos / LiDar data / Translation work to be labelled .
We will also be launching something called Human APP. Which will demonstrate our ML Marketplace & allow anyone to earn money by solving labelling tasks 🙂

Questions from Twitter

#1. We see your phrase “A New Way for Humans and Machines to Securely Connect and Collaborate”. Can you elaborate this?

Harjyot Singh: – That’s a good one. Our main focus since day 1 was to make work interesting for Humans again.

Since Human Protocol is an end to end toolset – it manages the job creation, the job matching to skilled workers, the solution verifications & micropayments – it had to be done in a truly permissionless manner which essentially means no party can influence any part of the process even us.

That led us to an even more interesting idea why should humans focus on the boring repetitive tasks when they can be done by automated software systems? To be Human is to channel creativity & analytical power & that’s how the protocol is designed.

#2. Please explain what is the advantage of HUMAN over similar projects? How can you capture the most value from this nascent market for data governance, among others? How will that benefit your tokenholders?

Harjyot Singh: – We are blockchain native, and that distinguishes us from the vast majority of similar services. We are also currently building out the future of blockchain 3.0 by creating a seamless L1/L2 interoperability while applications run on the top. That ecosystem is unique to HUMAN. This permissionless ness shouldn’t be taken for granted.

Would you rather trust the banks in a developing nation or software sytem to get you paid? 🙂

We also have immense domain experitise in the ML Space. We understand & have experience on both sides of the marketplace – the data requesters & data labellers (the latter of which we already have a network of millions!)
We also have plans quite soon to make sure that our future developments are done in line with the communities inputs. It is called the HUMAN protocol after all 🙂
So watch our social spaces for our comments on Governance & Various Grants / Bounty programmes.

#3. When they speak of “machines” in their documents, what did they refer to? Are there really robots that are part of the Human Protocol? How would the process of exchange of labor between a human and a machine take place?

Harjyot Singh: – Hah! Not yet – hopefully soon 😉 What we mean there are intelligent software systems. The interactions happen on many levels – on an infrastructural level it happens with machines really tailoring a job for an earner, verifying the solution quality & making sure the earners get paid instantly. A level higher might look like this:

Say the job is to draw boxes around cars & buses in an image some of these trained machines might “suggest their answer” & all the earner does is quality assure whether the machine’s answers are apt! Hope that answers your question 🙂

Questions from Telegram

#1. Why you decided to run your project in the BSC ? Will you support other blockchains too and expand more?

Harjyot Singh: – Our entire protocol has been desgined from the grounds up to be chain agnostic. Our intial implementation is on Ethereum, and we have also partnered with quite a few others – Solana, Zilliqa, Elrond, Skale to name a few. We believe that we can leverage benefits of all these L1s & L2s to deliver a much more polished solution.

#2. What triggered you to get started with the Human Protocol, and what kind of springs do you think there are with data in AI and ML? And how did you solve or solve these problems?

Harjyot Singh: – The team, the vision behind the project & the opportunity to be pioneers in building out technologies of the future.

I think ML Industries have a few problems:
– Data is too expensive since it’s controlled by a few parties.
– Data isn’t granular and diverse enough to build products that are really tailored to users.

This leads to good verified data not being available to anyone from researches, startups to even big companies. With our ML Marketplaces we will be able to bridge that data gap. We are already incorporating tools like Intel CVAT that *really* can create any media data you’d like.
Also since it’s a permissionless protocol we don’t get involved in the costs, that’s between the requester & the labeller 🙂

#3. Safety and security are always of the utmost importance. So, what are the security mechanisms of HUMAN to ensure that users’ assets are not targeted by hackers? 

Harjyot Singh: – All our software is carefully audited by the best in their business 3rd parties, these reports are available on our github. We also by our nature made a lot of the core tools we’ve built to be Open Source, which allows us to prioritise & constantly incorporate community solutions 🙂