The wild story behind four million YouTube videos that shape how AI sees, writes and creates
- Advertisement -
Related Posts
- Advertisement -
- The opt-in AI training of YouTube changes from makers to silent architects of future technical tools
- Many makers say yes to AI training access, even if no money is involved
- Oxylabs collected millions of videos in a data set that can trust AI developers ethically
An increasing number of YouTubers is able to train models with the help of their videos, and surprisingly, many do this without direct compensation.
Under YouTubeThe current setup, makers are given the option to register by checking compartments that give permission to around 18 large AI developers.
If no box has been selected, Youtube does not allow the use of that video for AI training purposes. This means that the standard posture is non-participation and that every admission is completely voluntary.
Makers choose an influence on income
The lack of payment may seem unusual and the motivation seems to depend on influence instead of income.
Makers who choose to perhaps see it as a strategic step to shape how generative AI tools interpret and present information interpret and present – by contributing their content, they make it more visibly in reactions generated by AI.
As a result, their work can shape how questions are answered by everything from Ai -writers Unpleasant Large language models (LLM) for coding.
Oxylabs has now launched the first YouTube data set based on permission, consisting of four million videos of a million different channels.
All contributors are explicitly agreed to use their content for AI training, and according to Oxylabs, these videos, complete with transcriptions and metadata, are carefully compiled to be particularly useful for training AI in tasks for generating images and videos.
“In the ecosystem that is aimed at finding a reasonable balance between respecting copyright and facilitating innovation, youtube flow line is giving permission for AI training and offering makers flexibility an important step forward,” said Julius černiaus greenhouse, CEO of Oxylabs.
This model not only simplifies the process for AI developers who are looking for ethically produced data, but also takes care of makers about the use of their work.
“Many channel owners have already registered to use their videos in developing the next generation of AI Tools. This enables us to make and offer high -quality, structured video scene sets. In the meantime, AI developers have no trouble verifying the legitimate origin of the data. “
However, broader concerns remain about how government organizations and legislative authorities tackle similar problems.
The Bill of the UK data (use and access), for example, got stuck in parliament, as a result of which figures such as Elton John criticize the government’s handling by the government.
In this legislative vacuum, makers and developers will probably be confronted with uncertainty.
Oxylabs presents itself as filling that gap with a model -based model, but critics will still wonder whether such initiatives tackle really deeper issues of value and fairness.
Maybe you like it too
- Advertisement -
