Microsoft Accelerator startup DefinedCrowd connects machine learning with native speakers
Historical source listing. Original English-language article published by TechCrunch on 2016-06-03. Read the original publisher for the full report.
What was reported?
Historical source listing. Original English-language article published by TechCrunch on 2016-06-03. Read the original publisher for the full report.
Context and implications
A few impressive demonstrations do not establish a model's reliability. Evaluation tasks, datasets, failure rates and real-world operating conditions must be examined independently.
The value of AI in practice depends on accuracy, cost, latency, privacy and effective human oversight. Independent testing should be distinguished from a product demonstration.
What to watch next
Further developments are best compared with the original publisher report, the underlying measurements and any follow-up experiments or product releases. The scope of this source is not the same as independent verification of every subsequent claim.
Illustrative news photo · Unsplash / Unsplash ↗
Read at original source ↗
Comments
Reader question: What is the biggest opportunity or risk of this AI development?
No comments yet. Be the first to contribute.
Reader discussionIllustrative examples
The names and comments below are illustrative, not real visitor posts. Likes and replies in this section remain on this page. Use the form below to post a real comment.
This sounds promising. What would be the most valuable real-world use case?
It is worth comparing the source announcement with independent measurements.
A practical example of this technology in a company would be useful.
✎ Write a comment