The speech-to-text project uses a network of audio ‘miners’ on the Waves network in an approach that is not only more cost-effective, but that presents a new economic paradigm.
Every now and then a new idea comes along that represents not just a quantitative improvement on the status quo, but something qualitatively different. Bitcoin itself was one such idea. It wasn’t as technically groundbreaking as it may seem; like most innovation, it built on existing and established ideas. What it did do was put together a number of different components in a way that gave rise to a totally new kind of system with new economic and technical properties — and one that has far-reaching social implications.
Distributed speech processing
Anryze is another such idea. The company, which has just launched its crowdsale, takes an entirely new approach to speech-to-text processing. Whilst this is not only more efficient than existing means with no loss of quality, it also offers a new paradigm of ownership.
Like bitcoin, Anryze doesn’t create a new paradigm from scratch — although its speech-recognition algorithms are new and, even in alpha, superior to those of the current market leaders. It puts together existing pieces to create a self-sustaining and self-contained ecosystem: a DAO, of sorts, that operates independently of a centralised company but that has the benefits of both worlds.
Speech-to-text services are currently dominated by a handful of large players such as Google and IBM. Due to overheads like employees, premises and data centre costs, plus the need to pay shareholders, these tend to be expensive. There is certainly room for disruption in this fast-growing sector.
Having developed its own speech recognition software, Anryze recognised the need to deploy its technology in a way that was more cost effective than the competition. The solution the team hit upon was to create a decentralised network of nodes that would process audio files on a peer-to-peer basis. It is an intriguing twist on bitcoin’s proof-of-work: instead of generating hashes, which are critical for the security of the network but otherwise all but useless, Anryze’s network members will undertake the computationally expensive work of processing audio files and returning them as text.
The fuel of this system is the RYZ token, which is hosted on the Waves platform. RYZ is currently being sold in the company’s crowdfund and will be purchased by clients from the open market to pay for speech-to-text services at a dollar equivalent amount. Audio ‘miners’ will receive 70% of RYZ paid by clients, with the remainder distributed to existing RYZ holders and to the team, to fund ongoing development. As the service becomes more widely used, growing demand for RYZ will increase its price.
The efficiencies of cutting out the middlemen are such that the costs of using Anryze’s speech recognition service are as little as a third as much as Google or IBM charge. But the model doesn’t just allow businesses to do the same thing for less. It allows them to participate in and take ownership of the services they use.
Anryze’s approach enables ordinary companies to deliver the same service they use and profit from others’ use of it, as well as paying less when they use it themselves. It is a new model of ownership where users become service providers and financial stakeholders, totally upending the existing norm of corporate ownership the pressure to maximise dividends for shareholders. It may not be the first initiative to do this, but it’s certainly one of the better thought-through and credible instances of a DAO.
The Anryze tokensale has reached its minimum target of $1.6 million. The crowdsale will end on 12 October or when the cap of $6 million is reached. For more information about Anryze, visit https://anryze.com.
Why the Anryze ICO is socially as well as technologically groundbreaking was originally published in Waves Platform on Medium, where people are continuing the conversation by highlighting and responding to this story.