Science starts with a question. Data science starts with the data. What makes data science so hard is that it starts in the wrong place. https://t.co/AQKbPCCtC2
— Roger D. Peng (@rdpeng) August 22, 2018
Science starts with a question. Data science starts with the data. What makes data science so hard is that it starts in the wrong place. https://t.co/AQKbPCCtC2
— Roger D. Peng (@rdpeng) August 22, 2018
Anyone who thinks vision has already largely been solved in AI is kidding themselves. Object recognition ≠ scene comprehension. https://t.co/hVIi6ly9yS
— Gary Marcus (@GaryMarcus) August 21, 2018
The scandal isn’t what’s retracted, the scandal is what’s not retracted. https://t.co/FwFMXsNikC
— Andrew Gelman (@StatModeling) August 21, 2018
We need a Goldilocks Rule for AI:
— Andrew Ng (@AndrewYNg) August 17, 2018
- Too optimistic: Deep learning gives us a clear path to AGI!
- Too pessimistic: DL has limitations, thus here's the AI winter!
- Just right: DL can’t do everything, but will improve countless lives & create massive economic growth.
Deep learning, in many cases, enables engineers to replace complicated, heuristics-laden machine learning pipelines with a simple end-to-end model (plus a data collection & annotation process, typically easy to maintain). https://t.co/J0VWz76xvf
— François Chollet (@fchollet) August 16, 2018
There is rarely "one weird trick" that will provide the definitive solution to a complex issue. But inversely, just because a problem looks difficult at first, don't assume it will necessarily require a complicated solution. Keep a bias towards simplicity.
— François Chollet (@fchollet) August 16, 2018
Thinking is better done in writing. And the language in which you write affects the scope of the thoughts you can think -- absence or presence of vocabulary for certain concepts, degree of precision of word nuances, etc. Until Cicero, no one would do philosophy in Latin.
— François Chollet (@fchollet) August 12, 2018
Too many people in the field of AI are chasing the latest fashions. My advice: keep your eyes on the fundamentals, focus on the long-term challenges. The important questions are still the same today as they were 20 years ago.
— François Chollet (@fchollet) August 11, 2018
"I have very strong opinions about programming, but one rule I try to follow is 'do not mock other programmers'.
— 👩💻 DynamicWebPaige @ $HOME 🏞 (@DynamicWebPaige) August 11, 2018
Programming is too big, and I’m too small to understand everything.
Disagreeing is fine; laying out why people are wrong is fine; making fun of them is *not* fine." pic.twitter.com/uBoiqc7rxS
the longer I hang around in academia, the more upsetting I'm finding its problems: the obsession with prestige (both personal and institution reputation), the exploitative pyramid employment structure, the systemic sexism and racism, the paper and grant app merry-go-round.
— Cian O'Donnell (@cian_neuro) August 10, 2018
The single best thing you can do when asking for coding help online is provide a short, complete example script that others can copy, paste, and run **without any modification** to reproduce your problem.
— Jake VanderPlas (@jakevdp) August 9, 2018
This applies to StackOverflow, mailing lists, github issues... everything.
Prediction 2: The currently hip ai algorithms that require endless samples from a simulation will never get us to generalizable ai capabilities (or agi).
— Richard (@RichardSocher) August 9, 2018
Some things can't be simulated/sampled until solved like natural language or many other important areas like medicine.