Deep Learning: What It Is & What It Can Do For You • Diogo Moitinho de Almeida • GOTO 2017
This presentation was recorded at GOTO Amsterdam 2017. #GOTOcon #GOTOams http://gotoams.nl Diogo Moitinho de Almeida - Research Engineer at Google Brain ABSTRACT We will start from the basics of deep learning: what it is, how it works, and how to get started, and then move to the most commonly used architectures including convolutional and recurrent networks. We will then talk about its capabilities [...] TIMECODES 0:00 Introduction 3:15 Gradient Descent 13:51 No: Unsupervised Learning 16:41 Yes: Image Classification 17:40 Not Yet: Unbiased Image Classification 22:58 No: Software Development 30:59 Not Yet: Game Playing 34:27 Yes: Face Transformation 35:40 Yes: Art From Scratch 39:10 Yes: Style Transfer 41:31 How Will The World Change? 43:08 What Can Deep Learning Do? 44:38 How Do I Take Advantage Of These Trends? Download slides and read the full abstract here: https://gotoams.nl/2017/sessions/94 https://twitter.com/GOTOcon https://www.linkedin.com/company/goto- https://www.instagram.com/goto_con https://www.facebook.com/GOTOConferences #DataScience #DeepLearning #ML #AI Looking for a unique learning experience? Attend the next GOTO conference near you! Get your ticket at https://gotopia.tech Sign up for updates and specials at https://gotopia.tech/newsletter SUBSCRIBE TO OUR CHANNEL - new videos posted almost daily. https://www.youtube.com/user/GotoConferences/?sub_confirmation=1
Read Video · Transkript & İçgörüler
Bu bölümün tam transkripti + AI içgörüleri var
Ücretsiz hesap · kart gerekmez · kayıtta 150 kredi, bu bölümü açmaya yeterli
- 📄 Zaman damgalı tam transkript
- ✨ AI özeti, anahtar kelimeler ve zihin haritası
- 💡 Ana çıkarımlar ve alıntılar
Bölüm zaman çizelgesi
Introduction, speaker background, and why deep learning matters now
- Diogo Moitinho de Almeida works at Google Brain and frames the talk around why the audience should care about deep learning given the current hype cycle.
- He establishes credibility by noting he broke a 13-year losing streak for the Philippines at the International Math Olympiad and won top prize in the International Modeling competition.
- Jeff Dean's quote that if you're not considering deep neural nets for your problems you almost certainly should be is used to motivate the audience.
What deep learning actually is: gradient descent and stacked simple functions
- The core of deep learning is framing a problem so that what you care about (or a proxy) is differentiable, then minimizing it via gradient descent.
- Deep learning is essentially machine learning made 'deep' by stacking many simple functions together, which yields something far more powerful than any single function.
- The hard part is knowing that the simple algorithm will still work for very complicated, non-convex models with stacks of layers.
Survey data on AI progress and the difficulty of predicting what deep learning can do
- A survey of AI researchers shows Angry Birds rated roughly as difficult as the World Series of Poker, and AI researcher rated significantly harder than math researcher with a roughly 50-year gap.
- Opinions diverge wildly on AGI timelines, with some researchers predicting 10-15 years and others over a hundred years; the speaker places himself in the latter camp.
- Andrew Ng's heuristic that any mental task a typical person can do in under one second can probably be automated is offered as a rough guide.
Temel kavramlar
- deep learning— The central topic of the talk, defined as stacked simple differentiable functions optimized by gradient descent.
- gradient descent— The single simple algorithm that underlies almost all recent machine learning progress.
- supervised learning— The speaker's recommended approach: label data and tell the model the correct answer directly.
Önemli alıntılar
if a typical person can do a mental task with less than one second of thought we can probably automate it using AI either now or in the near future
🔥— A concrete, memorable heuristic from Andrew Ng that reframes the question of what AI can automate around human reaction time.
how I feel about reinforcement learning is that it can work but you don't want to rely on it working
💡— Directly challenges the hype around deep reinforcement learning by warning that betting your company on it often ends in vaporware.
Uygulanabilir çıkarımlar
🧠Understanding Deep Learning
Deep learning is essentially stacking simple differentiable functions and minimizing them with gradient descent.
This week, write down one problem you care about and check whether a proxy of its objective is differentiable.
The hard part is knowing that the simple algorithm will work for very complicated stacked models.
Pick one popular architecture (e.g., ResNet) and read its paper to see how simple layers are stacked.
🎯Choosing the Right Approach
Supervised direct pattern matching is where deep learning reliably works; reinforcement learning is risky.
For your next ML project, spend one week labeling a small dataset instead of trying unsupervised or RL methods.
If a typical person can do a mental task in under one second, AI can probably automate it.
List three tasks at your job that take under one second of human thought and assess if they can be automated.
Transkript ve içgörüler yapay zeka tarafından oluşturulur ve hatalar içerebilir. Doğruluk, ses kalitesine ve konuşmacıların netliliğine bağlıdır — bir şey yanlış görünse orijinal ses her zaman doğru kaynaktır.
Bölümler ve videolar okunmaya hazır
Podcast bölümleri

Why you need the growth mindset of a weed | Stephen Nobles
TED Talks Daily
20 Ağu 202618:24EN
ChatGPT – The Super Assistant Era | BG2 Guest Interview
BG2Pod with Brad Gerstner and Bill Gurley
15 Mar 20261:03:40EN
Episode #218 ... Dostoevsky - Notes From Underground
Philosophize This!
17 Ara 202433:37EN
Kadınlar ve Erkekler Neden Arkadaş Olamaz?
Kendine İyi Davran
22 Eyl 202617:55TR
No.221 雷鸟 CEO:新技术越来越多,我们为什么还需要一副智能眼镜?
三五环
27 May 20261:27:43ZH
Folge 1 - Ein verschwundenes Land
Zeitreise DDR
9 Kas 202522:49DE
Videolar

We Tested an AI Agent That Builds 1000 Ads in 10 Minutes
Marketing Against the Grain
26 Şub 202635:44EN
Who’s Afraid of Chinese Models? | Stratechery by Ben Thompson
Stratechery
29 Tem 202620:25EN
Making $$$ with Grok Bot
Greg Isenberg
21 Ağu 202644:21EN
«Современный урок по ФГОС: требования, этапы, цифровые решения»
ЯКласс
11 Nis 20231:39:24RU
#29 Der Sachsenspiegel
99 mal Geschichte
22 Eki 202550:26DE