jasonlaska

5 ms·

1.
by jasonlaska 10y ago
Part 1 can be found here: https://news.ycombinator.com/item?id=12039000
2.
by jasonlaska 10y ago
We've posted a follow up on the humans behind Clara here: https://news.ycombinator.com/item?id=12074657
3.
by jasonlaska 10y ago
Our follow up on the humans behind is here: https://news.ycombinator.com/item?id=12074657
4.
by jasonlaska 10y ago
Even modern methods today (deep recurrent networks, etc.) can do pretty well with these kinds of tasks (a very large ontology for instance) if you have enough annotations of the nuance!
5.
by jasonlaska 10y ago
"The customer service test" -- I like it.
6.
by jasonlaska 10y ago
Your last comment on the seemingly circular relationship between Clara and a "turing-test-passing bot" is especially salient. This reference (found in the second footnote in the post) http://www.ijcai.org/Proceedi
7.
by jasonlaska 10y ago
This is noted, thanks!
8.
by jasonlaska 10y ago
Clara has a 1-hr SLA for the processing of an incoming message. While I cannot give numbers on the speed of annotators (or volume), I can say that our platform is designed to enable quick and accurate work via incentive mechanisms. We avoi
9.
by jasonlaska 10y ago
We think it's a good idea too ;) Machines and humans truly have different talents: machines are great at memory, keeping track of state, distributing information while people are great at understanding subtle nuance in natural langua
10.
by jasonlaska 10y ago
It's certainly possible. One advantage of our setup is that rather than getting ok -to- noisy labels from customers, our CRAs understand the end-goal of the application and generate pretty great data. We are also able to incentivize
11.
by jasonlaska 10y ago
Hi! I'm the author of the post and happy to answer any specific questions here.