‘Blockchain and AI have disrupted the old-fashioned idea that words need to have meanings. Two of the main things computers are good at are keeping records and executing algorithms. Everything in the former area is apparently “blockchain” and everything in the latter area is “AI”. So if you’re selling software it’s safe to add those words to your marketing copy.’
— Arvind Narayanan
If the above quote is anything to go by, it appears the holy grail of computer technology boils down to a collision of two of the most hyped technologies today: blockchain and artificial intelligence. TLDR, this article is about how Nebula AI got it figured out by merging the two technologies🙂 right? Nothing more to see here. Well, if that’s the first thought that crosses your mind… guilty as charged. Read on, this is not about throwing buzzwords around to create hollow excitement; it’s going to be fun as we discuss how Nebula AI intersects the two fields for maximum value creation. As the author I only get to say trust me on this one a limited number of times; this is one of those times. Now am getting carried away, back to business.
With both blockchain and AI still relatively in their infancy, I believe their unique synergy when merged together can unlock so much possibilities on top of their latent value. Blockchain automates contractual obligations and enables a secure, free, trustless interaction in P2P networks without any central oversight. Since Bitcoin redefined finance, blockchain interest has extended into other fields including health, data, democracy and many more. The marketcap of all cryptocurrencies today stands at $200 billion, just a fifth of Apple and Amazon’s valuation.
On its part, AI ushers in the augmentation age through:
✔️making computers smarter and
✔️making sense of data informing business decisions.
Autonomous vehicles, vision, healthcare, robotics, and IoT are some of the fields that are made possible by AI.
‘Artificial intelligence is the new electricity’
✅It is projected that by 2030, AI will add $15 trillion to the global economy and blockchain more than $3 trillion.
How blockchain and AI complement each other
Any AI model is only as effective as the amount of data used. AI models use datasets to ‘learn’ and choose the most optimal function. How to collect, transmit and avail this data for AI training is very crucial. Today, artificial intelligence is a field that is predominantly the reserve of centralized organizations such as Google, and Facebook because they have access to large computational power and large datasets. The monopoly of AI is dangerous because its value is commercialized for corporate advantage leading to asymmetry between value creation and corporate interest. Blockchain technology can make it possible to encrypt, store and share data for AI use without intermediaries. The interaction of AI models with the data store is managed by pre-coded protocols preventing any misuse. In summary, with blockchain AI can be decentralized.
☑️Fun fact: Following the 90s software trends of open source software gaining popularity over closed source counterparts, blockchain technologies, with decentralized communities and decentralized protocols have one less hurdle to deal with.
How can the augmentation age look like?
Currently, AI value is the reserve of a few corporations. Blockchain can be the catalyst to accelerate and democratise AI for common good. Take a simple use-case like object detection. While going about your business, unlike the past when you had to focus, with AI you can find a viable stand in to carry out mundane repetitive tasks like identify and update what has run out of stock. Think of facial recognition but now for everything like commodities in this case. If you want to check how many oranges are left, the AI can do it for you.
In a previous Autodesk University workshop, a speaker said that if a person was like a computer we would have fired them long ago. Of course, a computer has its strengths in terms of automation and computers have come a long way since their origins in the 20th century. However, despite being the critical tool of today’s workplace, computers are only as good as the person running them. We have to give a computer/smartphone the same instructions each and every time we use them. Think about how we can optimize workflow, save costs and time if not only the computers could be able to learn and repeat our most important tasks. Next time you use your computer and device but do not have to carry out the initial repetitive pattern but also the computer can also suggest some ideas like how Hollywood pictures intelligent robots. Billions of dollars in running costs can be cut as enterprises optimize their workflows.
Take the field of medicine. Today we have fitness apps that help us to achieve our fitness goals. With AI, we can take this further. For instance, an AI that can be able to recognize appropriate body shape when carrying out exercise can be used to incentivize proper workout through a blockchain token economy. Disease scanning can be improved orders of magnitude higher through AI. Researchers can be able to identify malignant body cells in an easier manner. Think of how many lives can be saved if cancer diagnosis improved from today’s methods. Add to that the ability of blockchain to properly record medical data, and medical field would be given a huge boost.
Fun times ahead!
Decentralized AI
All the above described values of AI, cannot be achieved with the current AI monopoly. Or if they are, then they will come at a very great cost to the common person like you and me. That is why we need decentralized AI.
For decentralized AI to be a reality, we need to chew a few vegetables first:
Can we find third parties who can contribute to training AI models❓
Privacy: training data be obtained without being privacy-intrusive❓
Economic incentives to third-party training parties❓
Can the AI model perform in transparency and trustlessness❓
One of the project which seeks to make decentralized AI a reality is Nebula AI.
NEBULA AI
Nebula AI is a 3rd generation blockchain project for the creation of autonomous decentralized AI. Decentralized AI apps otherwise known as DAI apps can be launched on Nebula AI and be accessible more cheaply by everyone than existing centralized AI projects.
Nebula AI solves one of the most pertinent problems in the AI space, the high computational load to run complex model training, by connecting a network of GPU owners with AI developers.
High computational power is very important because it enables more data to be processed. The more the data processed, the faster the AI learns and consequently, the better the final AI algorithms.
Recent breakthroughs such as cloud computing have reduced this deficit. However, much of the computing power is still in the hands of a select few tech giants. Nebula AI updates traditional GPU mining machines into AI computing devices. This means that AI developers no longer need to worry about setting up complex mining rigs to carry out the computation, they can plug in to the Nebula AI cloud engine at a small fee.
Mining
The network of GPU owners will act as miners. NBAI mining will occur through anyone contributing their computing power(graphics card) and earn NBAI tokens. By offering their computational power to carry out AI algorithms, miners will be able to earn NBAI tokens as a reward.Large tasks can be split and processed parallel speeding up the speed.
Mining will initially be POW until 2018 Q3, on the Helix chain but this will be updated to Proof of Group(2019 Q1 onwards) on Orion chain.
Economic models
Users can pay NBAI to use DAI apps. Developers can submit AI tasks to NBAI and pay using NBAI tokens.
DAI Apps
To support enterprise usage NBAI, the NBAI main chain will be scalable, flexible and secured. By 2020, more than 500 DAI apps will run on the NBAI mainnet.
Some of the DAI apps already developed on the platform include:
📌Quant AI — analyzing market movements AI
📌Sentiment Analysis — attitude expressed in sentiment, like in consumer behavior
📌Biomed AI — disease diagnosis
By the end of 2018, NBAI will support 10 DAI apps. By the end of 2019, that number will stand at 500.
Mainnet|Miner Recruitment
On September 28, 2018 Nebula AI will launch their mainnet. Currently, the first round of miner recruitment is taking place. The building blocks of the Nebula AI network are:
📌Graphics threshold are at least GeForce GTX 1060
📌I3, i5, i7 etc
📌Hard drive more than one TB
📌Memory at least 16 GB
📌System Ubuntu 16.04 LTS
📌Network bandwidth 50 Mbps download speed and above
You can register here to be part of the Nebula AI mining community
https://docs.google.com/forms/d/e/1FAIpQLSfeqIThls7RRFRNEdmvVO2wXOChzfdpeBebAaWo6gcu1DQPng/viewform
My parting shot to Nebula AI team — Carpe diem!
I liken the race for AI for the common good with the contribution that the Wright brothers made for flight. It is not the huge corporations then, similar to today’s tech giants, who made flight breakthroughs despite their large financial backing. Rather, that proud recognition is for the Wright brothers, who were driven by their passion to make the world better. They say history doesn’t repeat but it rhymes and all waves were originally small waves, barely worthy of notice. I believe now is the time for Nebula AI to write the next chapter of the humankind story like pioneers such as the Albert Einstein, the Wright brothers and Neil Armstrong beforehand. Steam on!
For more information on Nebula AI:
Website: https://nbai.io/ Linkedin page: linkedin.com/company/nebula-ai/
Whitepaper: https://www.nebula-ai.com/whitepaper/NBAI_whitepaper_EN.pdf
Bitcointalk ANN: https://bitcointalk.org/index.php?topic=2928531.0
Telegram: https://t.me/NebulaAICommunity
Twitter: https://twitter.com/nebula_ai
Github: https://github.com/nebulaai
Bitcointalk username: Marahunter
For more information, follow the following links:
https://medium.com/datadriveninvestor/why-decentralized-ai-matters-part-iii-technologies-930c3c9d10d
https://medium.com/datadriveninvestor/why-decentralized-ai-matters-part-i-economics-and-enablers-5576aeeb43d1
https://medium.com/datadriveninvestor/why-decentralized-ai-matters-part-ii-technological-enablers-a67e3115312e
