Can An AI-Based Marketplace Streamline the Path to Deep AI?
Deep AI or General AI has been one of the biggest fascinations of much of the tech world.
Many kinds of research are being carried out to achieve it. This is simply because it is the highest level that the Artificial Intelligence (AI) technology can achieve. In fact, it is the closest that technology can get towards human intelligence. This is to say it will be able to almost do all the mental tasks that humans are able to perform.
There are both people who are welcoming of the idea of a deep AI, and those who are apprehensive about it. Not only scientists but also philosophers and social thinkers are concerned about the fate of the world in the presence of these super robots. However, in reality, we are far from achieving this.
The present AI technologies available to us are those which are called narrow AI. These are suited for performing very specific tasks. In fact, while the human brain can perform about one billion-billion calculations per second, the most powerful computer in the world has a long time to catch up.
Tianhe-2 developed by the National University of Defence Technology of China, which holds this record, now can only carry out 33.86 quadrillions of calculations per second. The greatest achievement of scientists, in the process of emulation of a brain, has been the emulation of a 1-millimeter flatworm consisting of 302 neurons.
Build better voice apps. Get more articles & interviews from voice technology experts at voicetechpodcast.com
Machine Learning and Deep Learning
Machine learning is crucial for the process of the development of any AI system. While for narrow AI systems, any machine learning system based on specified labeled data sets may be enough, for the deep AI system a special kind of machine learning is required, which is being called deep learning.
This is because deep learning techniques are more robust and rigorous and do not require labeled data. This is to say, that the deep learning process is much more autonomous of human interference compared to other standardized machine learning techniques. The deep learning systems operate through neural networks, which are developed based on the prototypes of brains. The essential thing in these, therefore, is the ability to draw upon patterns autonomously.
Unlike normal machine learning systems, which are developed by the use of large data sets used for their training, neural networks also involve inference for their development. Inference refers to the capacity of the system to autonomously identify the kind of sources of data that will be required for its development. It is here that the question of an integrated platform becomes the most relevant. OLPORTAL can be such a platform which will be important for the progress towards deep AI.
The Need for an AI Marketplace like OLPORTAL
As we see data becomes a key element in the development of neural networks. While data is mined in large proportions by many agencies and commercial organizations, very few companies are able to use such data for the development of their neural networks. Given the decentralized nature of the OLPORTAL, it can become an immense repository of meta-data, which will be capable of training the neural networks. Using meta-data will also ensure the privacy of users.
On such a platform, models can be developed to allow for pulling off the computing power of the machines of the various members of the community.
All these different features help in the development of more sophisticated neural networks. This is not possible in the case of centralized networks as they have very limited access to users’ data. Since the users do not have any stake in their research, they have no incentive to share their data. However, in the case of a decentralized platform data sharing is an inclusive process, where all the stakeholders can benefit equally.
Parting Words
OLPORTAL has different kinds of neurobots already on the platform. Therefore, their development can also be improved through the repository of meta-data that can be generated within the platform, giving further incentives for the developers to push for the collection of certain kinds of data.
Only through such an inclusive system can a large scale task of attaining the goal of deep AI is possible. The various steps need to be incentivized, which is only possible through a market mechanism like the one provided by OLPORTAL.
To understand what we are building at OLPORTAL, please connect with us –
Website | Telegram | Twitter | Medium | Facebook | Reddit | Linkedin | Instagram | Kakao Talk
