Profile picture
Russell Kaplan @russelljkaplan
, 3 tweets, 1 min read Read on Twitter
What makes large companies have a strategic advantage in machine learning research & products?

Answer for the last five years: data
Next five years: compute

More & more DL research is becoming inaccessible without Google-scale datacenters.
Custom AI chips will give a temporary free lunch for the masses seeking much-needed addtl compute. Will be similar order of magnitude speedup as CPU->GPU transition. But in the long term, many research & product breakthroughs will be inaccessible to most orgs.
It seems like everyone in my lab has a part time gig at Google/FB/DeepMind/MS just for the extra compute. And that’s at Stanford, where AI funding is not exactly lacking
Missing some Tweet in this thread?
You can try to force a refresh.

Like this thread? Get email updates or save it to PDF!

Subscribe to Russell Kaplan
Profile picture

Get real-time email alerts when new unrolls are available from this author!

This content may be removed anytime!

Twitter may remove this content at anytime, convert it as a PDF, save and print for later use!

Try unrolling a thread yourself!

how to unroll video

1) Follow Thread Reader App on Twitter so you can easily mention us!

2) Go to a Twitter thread (series of Tweets by the same owner) and mention us with a keyword "unroll" @threadreaderapp unroll

You can practice here first or read more on our help page!

Related threads

Profile picture
✨🧠 The ecosystem that has grown up around @TensorFlow in the last few years blows my mind. There's just so much functionality, compared to some of the other, newer frameworks.

👉Consider this an ever-expanding thread for me to take notes + wrap my brain around products. Ready?
1) @TensorFlow Extended (TFX)

It's no secret that I 💕 #TFX and all of its tooling for deploying machine learning models into production. If you care about keeping your models up-to-date and monitoring them, you should check out the product + its paper.

tensorflow.org/tfx/?hl=zh-cn
2) @TensorFlow Hub

If you want to train your model on a small data set, or improve generalization, you'll need to use something called transfer learning. #TFHub modules make it easy—and are available in an #OSS marketplace: tfhub.dev.

site: tensorflow.org/hub/
Read 39 tweets
Profile picture
Kicking off day 1 of #ODSCWest. I'll be tweeting through the next few days and pinging #GCDigital folks for interest or potential follow-up. #opensource #DataScience #openeducation @School_GC
First up: @asuonline will be hosting an educational analytics conference in March 2019 focusing on what we can learn from learner data and how to structure learning to better meet learner needs. #ODSCWest
#AI differential - being able to tailor products to specific users == better able to predict value to and from users. E.g. cost to deliver value to user vs. cost to company to deliver that value. Allows ruthless competitive focus - @DataRobot #ODSCWest
Read 21 tweets
Profile picture
Next up at NBER DEV: Natalia Rigol on "Targeting High Ability Entrepreneurs Using Community Information: Mechanism Design in the Field"
With Reshmaan Hussam and Ben Roth:
papers.nber.org/conf_papers/f1…
This paper is a nice application of some of the concepts described by Esther Duflo in her Master Lecture yesterday, combining field work, causal inference and machine learning.
Read 13 tweets
Profile picture
!!!NEW Q DROP!!! 06/13/18 Busy Day at the Castle!! Q posted @POTUS_Schedule from this morning along with two previous Q drops!! Thank you, President Trump for the #QProof !! @formerlydoxed caught this Proof this morning!! See here: #QAnon @realDonaldTrump
!!!New Q Drop!!! 06/13 12:26 Q replied to his previous drop with this: Why are many Iranian resistance (Freedom Fighters) tagging #QAnon on Twitter?
Big things coming folks!! Remember, Q said Iran was next!! #IranFreedimFighters @realDonaldTrump Previous Q’s re. Iran below!
!!New Q!! 14:30 Q says to slowly & carefully re_read this article that he’s posted 3x now!! breitbart.com/big-government… Read it 10x 50x 100x stopping after each sentence to analyze & understand the significance!! Their only hope is to win in Nov.!! #QAnon @realDonaldTrump
Read 58 tweets
Profile picture
While I agree in many ways with Shoshana and Paul on the Sisyphean nature of preventing all foreign influence, their analysis of FB's new requirements falls into the same trap as Congress. They are more concerned about the very existence of abuse instead of prevalence and impact.
In many ways, the tech and media/policy worlds are talking past one another by using completely different criteria on judging the abuse of tech platforms.
This is reflected in the debate over whether FB/Twitter are "media companies". Implicit in such a classification is the idea that a single instance of something bad being distributed is absolutely the fault of the distributor, as we would judge the NYT printing an untruth.
Read 18 tweets
Profile picture
It seems like every company is tripping over themselves in a rush to say their software is “powered by AI.”

But saying “powered by AI” is like saying you’re “powered by the internet” or “powered by computer code." By itself, it means nothing.

Here’s how I think about it:
1) Most “now powered by AI!” is just a rebranding of the same heuristics and rules engines software has used for decades.

When your email program keeps track of what folders you use most often and magically offers to file mails to them, that’s a simple algorithm, not “AI.”
Software has been adapting to users for decades.

The old Microsoft Office “smart” menus and toolbars rearranged based on your usage (“powered by AI!”) as early as 1999.

Learning spam and antivirus filters have been around for more than 20 years.
Read 15 tweets

Trending hashtags

#Italian #Imagine #NULS #data #amendment #quote #AI #Strauss #abortion #bias #million #StephenCullinan #FDA #exhaustion #ODSCWestt #chains #law #datascience #Kandinsky #dataviz #DataScience #computer #LISA18 #UXA18 #FullArmor

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just three indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member and get exclusive features!

Premium member ($3.00/month or $30.00/year)

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!