I understand the argument -& no doubt this is true in some cases & perhaps some entire fields- but it isn't true more broadly.— Chris Chambers (@chrisdc77) March 14, 2019
Of the ~150 Registered Reports I've edited, I'd estimate that only 1 in 10 have needed to adjust their analysis plans after seeing the data. https://t.co/3oNPuNVTTZ
I taught a lot of stats to v. sharp grad students & post docs from other depts. And, hands down, the most prevalent false hope was that by learning more stats they could 1. know exactly which analytical approach was "right" and 2. eliminate uncertainty/error in their results.— Jenny Bryan (@JennyBryan) March 5, 2019
Controversial opinion: it is easy to pull together a high level list of ought to’s and we are drowning in them. Let’s stop being self congratulatory for “thought leadership” and start doing things. It is much much easier said than done. https://t.co/A22exrMqOY— Rumman Chowdhury (@ruchowdh) March 4, 2019
Dear OpenAI: Please Open Source Your Language Model— Sebastian Ruder (@seb_ruder) February 19, 2019
Nice piece by @gradientpub's @hughbzhang that makes the case that "withholding the full GPT-2 model is both unnecessary for safety reasons and detrimental to future progress in AI". https://t.co/v0BB3bYdKf
ML is more easily accessible than ever before. But with great power comes great responsibility.— Denny Britz (@dennybritz) February 17, 2019
The real danger of ML is that it’s now easier to justify whatever you like by pointing to the output of an ML model w/o understanding the big assumptions underlying your experiments.
The causality of working hard goes the other way than I think people think about it. Clear alignment, growth trajectory, and good morale makes people work hard. Trying to affect the outcome directly is not very effective.— Erik Bernhardsson (@fulhack) February 15, 2019
It’s mind-blowing how little the human condition has changed over the past two thousand years, despite all this technology. We strive for and worry about fundamentally the same things. So when you see something new promising radical change - be skeptical.— Denny Britz (@dennybritz) February 15, 2019
As everyone has a different point of view, it's just collisions everywhere :S— Smerity (@Smerity) February 15, 2019
- When does a model go from "safe" to "dual use"?
- How much of a "dual use" delay do we need to add?
- Should we release to journalists first or researchers?
- How can small labs participate in PR?