Ze.Fi & Braincities: AMA Recap

Noctis Caelum
Ze.Fi | زي.فاي
11 min readNov 25, 2020

--

Welcome everyone to the #ZeFi AMA with Braincities!

So, today we welcome Braincities here for an AMA. With a name like Braincities, you know it’s a project that aims to be more than just skin-deep. AI has long been a tech buzzword, blockchain has long been a tech buzzword, but apart from a couple of dead 2017 ICO examples, we haven’t really seen any attempts to merge the two and bring them beyond mere buzzwords into meaningful products. Braincities are hoping to change that, and leverage these two technologies towards creating a smarter (and better) society. Those are some very big aims, so let’s find out from the team how they plan to make it happen.
Welcome Bibin (@bibinkp), Kazé A. ONGUENE, Kheire (@Kheirie_ha), Esso (@Essoman)!

First thing’s first, could you please describe Braincities to the ZeCrowd in your own words?

Kazé:

My name is Kazé A. ONGUENE. Designer of the first decentralised Infrastructure as a service for the Internet of Knowledge.

I am the founder and Chief Executive Officer of BRAINCITIES LAB. Futuristic Thinker and Entrepreneur with over 15 years experience in IT project management, recruitment, team building and web development, I studied Software engineering at ECE Paris and computer science at la Sorbonne. During the last 15 years, I used my creativity and resources to help companies like France Telecom, Pierre & Vacances and Le Figaro adapt their businesses to digitization and develop web-based projects.

In 2013 I decided to focus on a social and engineering challenge: Using AI, robots, and Blockchain to design the city of the future and a sustainable society. This ambition is the current mission of BRAINCITIES LAB.

I Founded the company in 2013. With a strong focus on human behaviour, skills and knowledge assessment.

You can consider BRAINCITIES LAB like a Deep Tech Company offering On-Premise and Cloud based solutions for AI Augmented Autonomous Maintenance , human capital management. We use AI and Dynamic Knowledge Bases to help decision maker to make the right decision at the right time. doing so we leverageour experience in both time series analysis and machine learning to augment predictive and autonomous maintenance for smart cities like Den Haag (NL), and private organizations like ENEL.

ZeTeam: That’s amazing! It seems as though technology research and development, while onboarding partners in the traditional research space with resources is going to prove useful for Braincities to achieve its mission. What is your partner onboarding process? What companies are helping to support your research now and in what capacity?

We started with the research center from Pantheon Sorbonne back in 2014, since we didn’t stop growing our ecosystem building technical and strategic partnership with leading organisation like: HPE, IBM, SAP.
Building a strong global ecosystem has a lot to do to our longevity and successes.
2013, even before the ethereum birth. Tell us a bit more about the intersection you envisioned, between AI and blockchain. For some people it is mostly “trading AI bots”, what else is there?

Well, I believe the AI space is moving in a direction that will soon see the generalisation of transhumanism. Not only people want to use the technology, the also want to merge with it.
for several reasons:

1. Because it is very convenient for global leader to make the inventory of every humans

2. Because there is a marketing of transhumanism (Elon Musk brain chip)

3. Because We are entering in an area where we will be competing not only with other human for the job, but we will also have to compete with robots and AIs

Is there not an issue that in order for your solutions to work and benefit from transhumanism, you have to wait several years, if not even longer, for transhumanism to become an everyday reality for a critical mass of people?

“The current version of our solution is a heterogeneous data refinery powered by a data preprocessing engine enabling the prediction of time series with over 95% accuracy.”
I’m curious what concrete steps will you take to ensure that is as close to 100% accuracy as possible?

Here you are talking about our 2025 goal.

But before we go there it is very important to understand why we are building this technology and what is the value we are trying to bring for the human community.
I personally believe people should have the choice.

In 2018 Porsche Asked us to help their people working on the production line evolve from there current job to new jobs within their organization.
We created PICNETRIQ our human capital development product designed for internal mobility and career coaching
the underlying idea was and still to create a Human Supporting Artificial intelligence.
This AI is different as it is not design to replace, nor compete or augmente humans.
This AIs are designed to complement Human to collaborate in a symbiotic way with them.
This is where we started. We started by endowing the system with Judging and Empathy capabilities.

What is the purpose to release a tradable token in the near future? What is your development timeline?


we create a token with 2 purposes:

Machine to machine transactions

Project support
This month we initiated a smart city project with the city of Den Haag.
We came up with what we call an “Eco systemic Autonomous Maintenance” platform

Maybe I can show them the diagram of our solution “Ecosystemic Autonomous Mantenance “ we made for the Living Lab in Scheveningen, Den Haag

Hello, yes sure. The main idea of “Eco systemic Autonomous Maintenance” is to be able to allow machines to self diagnose themselves, and to be able to detect that a certain fault or irregularity. The machine will then contact automatically contact responsible stakeholders to initiate smart contracts.

ok now when i hear machine to machine transactions, my first thought is to run and hide. Can you elaborate on what this means exactly?

It mean a public asset is able to make self diagnosis, contact a stakeholders (decision maker of the city of Den Haag, publish a call for tender with the referenced contractors in the city, authorize the payment, book a meeting with a technician)

Shall we fear autonomy that this project is creating, for AI machines? I still can’t get my head around the idea

no we should not, because the AI has its own wallet, it interaction with various ecosystem requires the validation of its request the wallet of the various stakholder (Human, entities and or Public and private assets).

We adopted a federative Data architecture to make sure it would be easy to isolate entities with abnormal behaviors
-> Depends if they have been programmed to obey Asimov’s 3 Laws or not

Azimov’ 3 law would not survive a self conscious autonomous system.

Creating an environment that is diverse / heterogeneous is the solution.

I mean the grid used to build the AI A should be different from the one use to write/create AI B.
To avoid monsters, and deprecation of its source code nature created diversity.
Also, if it would be possible to explain the need for a token?

Hello sure, the main goal of our contract is to be able to execute and record some actions, when predefined conditions are met.

In the case of a maintenance of a city a machine can order for a maintenance, record that order, and execute a token based order.

Well, let me ask this again for our AI fans and those that don’t really understand how your AI knowledge base filter out the “bias-free” data?

What makes it stand out from the rest of the “predictions” and how can you make sure that the AI won’t be continuously fed with biased data? (Perhaps @Kheirie_ha) could chip in :D

I noticed I have not answered this question; the flow of questions are very fast 🙂 Mostly, to make sure that an AI is not biased is by being careful with training data. We make sure that the training data is a true reflection of reality and that it does not hold any prejudice; thus it is important to carefully examine the historical data (e.g. check the sources they came from, the time slots etc.).

Does this mean the AI will retain some empathy from your guys’ original tests?

In the Den Haad use case we presented the Ai interact with Human using Natural Language processing. In this case voice is the Human Machine Interface

This is how the AI can mine the human personality, sentiment and adjust its own level of “empathy”
If I can rap it up in few sentences: we aim to provide an operating system configured to augment human cognitive and decision making capabilities with no need for intrusive surgery. The envisionned operating system comprises a memory structure, including a symbol allocation DB, and it integrates semiotic technology able to undrstand multidimensional concepts and correlate them with specific contexts. it leverage key compoent of the datachain foundation like the datawallet to simulate short term memory (STM), and compose experiences using residual and/or environmental data, It also use dynamic ontology and time series to generate dynamical Knowledge bases to predict the steady states of evolving processes Distributed ledger technologies are used to monitor the decision process of the AI (a way to make AI Auditable) and to notarize the transaction made within an ecosystem like Den Haad, Paris, Dubai or any large organization like Volkswagen

What incentive (if any) is there for people to be token holders when AI and Smart City tech relies so heavily on machine-to-machine communications

The volume of transaction, wil drive the value of the token up. When it will be largely adopted this might mean Peta octets of data per months, millions of decisions per day, and hundreds of thousands smart contract per month
being able to allocate tokens to the right community (ies) or project(s) is key
The token will power the community to participate in the project governance.
this is why DC stands for Decentralised Control



— — -

ZeCommunity Time!

Are you planning to promote your project in different countries, where English is not good? Do you have a local community for them to better understand your project?


To answer your question:

Yes, We have our Datachain Foundation Meetup community (4500+ people), which includes community members from all over the world. And we organize meetups, hackathon, masterclass, AMA, podcast, webinars etc with our community.

We believe in glocalization and our solution is open to all the countries.

Will be sharing all the social links at the end of the AMA, so that everyone can have a look at it

How many actual smart cities will there be in 2025 at the level where

A) it would make sense to use your token

B) they would be physically able to use your token

We have project with 3 major cities/area in the world:

Den Haag (Initiating)

Dubai (PoC delivered — project pending due to Covid 19)

Paris eastern area (HR project at the scale of the region)

By 2025 I believe there will be over 200 real smartcities in the world (ref: Dubai, Singapor, Neom city …)

Others wille be what we call augmented cities old cities like Paris , using Data, AI and the Smart grid to provide digital services to their people.

Can you talk a bit more about the token sale? When will it happen? The utility of the token is still unclear to me tbh


The Token grant access to data contained in Datawallets or Digital services provided by AI

Token Sale, We are currently preparing for the listing of the Token On ZBX.one We will share the details of the Token sale Kickoff with @zendetta team (within the next 2 months) and accelerate the growth of the project
How will interest be calculated for lenders on the Braincities platform? On what basis and how will the amount of funds for borrowing be calculated?

Hi Atia, not sure to follow you. are you talking about one of the ongoing PoC we have with VC fund who want to tokenise its Portfolio using the DC Token?

Hello everyone

I’m Muhammad Fawaz, Radiology doctor as a hobby maybe, and math is my main interest!

1. In the recent years many AI Neural networks achieved tremendous results in Games like Go (a much more complicated board game than chess. GO is not a game that responds to brute force calculation, rather it requires intuition and an instinctive appreciation of positions and beauty). DeepMind’s Alpha Go beat the world champion at that time Lee Sedol, a player who was arguably the GOAT was crushed by 4–1 match result. In their 1st game, Alpha Go made a move (now famous by the name move 37) which was considered a tremendous blunder to the eyes of any Go player, yet it won the game later proving the opposite. On the other hand, in the game Alpha lost, it made another weird move that actually was a blunder that Lee managed to exploit and take it all the way to a win. Alpha’s developers said at that time that correcting such mistake would be harder than teaching Alpha how to play Go! My point is, who’s capable to judge judging if an AI is taking the right decision or not? To what extent are you willing to accept wild decisions?

2. Back on AlphaGo. After the match with Lee, Alpha Go developers thought about letting the neural network learn playing go from the very start on it’s own and created what’s called AlphaZero Go. After it was done learning, AlphaZero Go was more than 100 times stronger than AlphaGo (the one that beaten Lee and its learning was influenced by human knowledge of go).. They played a 100 games match together and guess what! AlphaZero beaten AlphaGo 100 to 0 :) . The point is, how would you be sure about any and every kind of information you influence your AI with?

Sorry for talking a lot about games, but what’s life other than a game without Ctrl-Z option to undo the designs that we regret, also we all know what a tremendous effect the game theory had over economy!

Hello, this is a very interesting question. In our solutions we are still providing human validation. For example the AI would provide the top solutions for a problem and the human will validate. Then this validation is used as a training data. Also, every decision that the AI makes is stored in data wallets.

How will interest be calculated for lenders on the Braincities platform? On what basis and how will the amount of funds for borrowing be calculated?

I can not reply to this question. It would be a very interesting one.

We are designing an AI that will be in charge of the control of the inflation of the DC Token… We should be able to improve lenders experience using AI, Dynamic Knowledge bases, Psuychometry and Blockchain technology. But it is not a use case we are currently addressing.

Also, since crypto tokens are driven by speculation and trading on dexes, isn’t there a risk that most tokens could find their ways into the hands of speculators, not people who actually live in the city and have a vested interest in using the tokens properly?

We have designed a method. Token owner can only trade a portion of their Tokens for example 5% per quarter. Obviously this rate is not fix it would evolve depending of the demand, performance, token allocation. The AI in charge of the management of the inflation rate is in charge of this.

Token allocation to communities and projects is what is incentivised.


Well, that was it!
We would like to thank Braincities team for the very AI-based educational AMA, to follow up with their latest news and updates, please follow the links below:

Braincities Website:
www.braincities.co (http://www.braincities.co/)

Telegram channel: https://t.me/datachain

Contact Email: contact@braincities.co

Datachain Website: https://datachain.one/

Twitter: https://twitter.com/braincities

LinkedIn: https://www.linkedin.com/company/braincitieslab/

Facebook: https://www.facebook.com/BRAINCITIES/

Meetup: https://www.meetup.com/datachaindc/

Medium: https://medium.com/@BRAINCITIES

--

--

No responses yet