pesenti

9 ms·

31.
by pesenti 5y ago
That’s a good question and I don’t think we have all the answers yet. But I expect that many functionalities will need to request and get consent from 3rd parties.
32.
by pesenti 5y ago
No but they should: https://www.ftc.gov/news-events/press-releases/2019/07/ftc-i...
33.
by pesenti 5y ago
I encourage you to read the details of the news, in particular how the data was collected with full consent from all (something that most research datasets don’t do well). It doesn’t answer your question directly but gives you an idea on ho
34.
by pesenti 5y ago
I support Facebook AI. The actual news is that we are releasing an extensive egocentric dataset with associated tasks for the research community. Collecting such dataset in a private and responsible way is a significant challenge. Happy to
35.
Textless NLP(ai.facebook.com)
1 points by pesenti 5y ago|0 comments
36.
AI now enables robots to adapt rapidly to changing real world conditions(ai.facebook.com)
2 points by pesenti 5y ago|0 comments
37.
Facebook Goes Boring(nytimes.com)
2 points by pesenti 5y ago|0 comments
38.
by pesenti 5y ago
Our main usages of speech rec are: - Understanding for videos for content moderation, ranking and recommendations - Captioning - Assistant tech in Portal, Oculus and upcoming glasses
39.
Facebook’s New AI Teaches Itself to See with Less Human Help(wired.com)
2 points by pesenti 6y ago|0 comments
40.
Self-supervised learning: The dark matter of intelligence(ai.facebook.com)
3 points by pesenti 6y ago|0 comments
41.
What Happened with News on Facebook in Australia(about.fb.com)
4 points by pesenti 6y ago|2 comments
42.
by pesenti 6y ago
The communication around this change was very poorly handled. This policy update does not affect the privacy of messages with friends or family in any way: https://faq.whatsapp.com/general/security-and-privacy/answ
43.
WhatsApp policy change doesn’t affect non-business communications(twitter.com)
3 points by pesenti 6y ago|0 comments
44.
No Looking Back(tech.fb.com)
2 points by pesenti 6y ago|3 comments
45.
by pesenti 6y ago
Facebook employee here. You can definitely continue talking about any company business or decisions, even if they have political relevance, as you did before. This change only applies to discussions completely unrelated to work.
46.
by pesenti 6y ago
Similar work done at Facebook: https://engineering.fb.com/developer-tools/predictive-test-s...
47.
by pesenti 6y ago
I am an immigrant in the US. My family uses FB products (FB App, WhatsApp, Portal) to stay in touch across the world. It really improves our lives. It doesn't mean I believe everything that's done on our products is good for the w
48.
by pesenti 6y ago
I have answered part of your question before: https://news.ycombinator.com/item?id=19036507 Happy to answer more.
49.
Introducing Manage Activity(about.fb.com)
2 points by pesenti 6y ago|0 comments
50.
by pesenti 6y ago
I am curious, given the international nature of Facebook which makes it hard to rely on country authorities to settle content moderation disputes: what would be a better alternative? (Disclaimer: I work at FB)
51.
by pesenti 6y ago
We need robots who look like Chris Atkeson, he is more personable than Santa.
52.
by pesenti 6y ago
I have answered similar questions on HN before. See https://news.ycombinator.com/item?id=19036507 for example.
53.
by pesenti 6y ago
It does work pretty well with changes of topic (often integrating the broader context as long as it fits into a few turns).
54.
by pesenti 6y ago
The bot is really optimized on engagingness rather than humanness - in particular this is how we chose the hyper parameters. We did evaluate Meena vs. Human (31 vs. 69) and Meena vs. BlenderBot (35 vs. 65). But didn't do BlenderBot vs.
55.
by pesenti 6y ago
The team put a lot of effort on dialog safety. You can find more info here: https://parl.ai/projects/dialogue_safety/ and the paper here: https://arxiv.org/abs/1908.06083 . Still, unlike Tay,
56.
by pesenti 6y ago
We are evaluating the possibility. But this is really for research only. To go beyond that we need further improvements to the safety layer.
57.
by pesenti 6y ago
That's definitely a fair concern. We believe that open science and transparency are the right approach here. By releasing it, we ensure that everyone is on the same page with respect to abilities and defense.
58.
by pesenti 6y ago
Blog post: https://ai.facebook.com/blog/state-of-the-art-open-source-ch... Paper: https://arxiv.org/pdf/2004.13637.pdf Open Source: https://parl.ai/projects/recipes/ Ask
59.
Data for Good: New Tools to Help Health Researchers Track and Combat Covid-19(about.fb.com)
1 points by pesenti 7y ago|0 comments
60.
Self-supervised learning is the future of AI(thenextweb.com)
1 points by pesenti 7y ago|0 comments

More ›