Earlier this week, Radha Basu gave a keynote at the Future of AI conference in Tel Aviv. The keynote conveyed the opportunity and the important role that humans play in the bringing AI to life. The following is a brief summary of the presentation.
Follow along with the slides:
- AI Technology needs massive amounts of training data: average volume 1,000,000+ for most production machine learning models
- Not only do you need a lot of data, it needs to be “On-target” data with relevant input parameters that are useful for the technology’s performance
“70% of complicated analytical jobs is probably about the organizing and cleaning of data.” – Guru Banavar, former Chief Science Officer of Cognitive Computing, IBM (Watson)
As data increases in complexity and AI gets into more-multidimensional scenarios, the way training data is prepared, the precision and accuracy it needs at a pixel level also evolves in complexity. At that point, it is not just about human labeling (which is what it used to be), but about human intelligence, expertise, and judgment.
- To bring AI technologies to life, you need a scalable, agile team that is custom-skilled in AI Data Labeling tasks and has the experience to get it done effectively and efficiently.
- Combination of human judgment, tools, and training
- More “on-target” training data with insights from previous sets
- Enterprise approach brings process learning, process insights, iterative consistency, repeatability, and security.
This is just a subset of the services offered for AI services.
- After labeling over 17Million images and processing over 1.3Million human judgment text annotations, we have the expertise to offer insights into the process of creating on-target datasets.
- It’s not all about volumes – it’s about producing those volumes at a very high rate of accuracy
- AI Services is creating jobs in Artificial Intelligence across the world, offering opportunity where it previously did not exist.
- AI Services is all about the people and offering market-aligned custom skilling
- It is a NEW workforce that is Digital First (born digital)
- They are young – aged 18-33
- It is female dominated with over 55% of our workforce is female – one of our delivery centers is all women. Their community is known for producing textiles, the level of detail and accuracy needed to do this work is innate to them and it shows in their work.
- Over 80% of AI Services workforce comes from a Social Impact background.
AI Services Bring AI to Life With Human-Empowered Computing