Cassie Kozyrkov (Google) on Decision intelligence (ML++) | TNW Conference 2018

Interested in coming to TNW 2023 🎡 ? Look no further, you can register today thenextweb.com/conference
Cassie Kozyrkov (Google) on Decision intelligence (ML++) at TNW Conference 2018
What is machine learning? What is AI? It turns out that these words have multiple meanings... and when businesses don’t know the differences, they run into a world of problems. We’ll go back to basics and strip away the jargon to see what’s really behind the magician’s stand in simple analogies. Let’s take a look at what’s easy, what’s hard, how to spot opportunities, and what you need to know to avoid the two biggest threats in AI. Along the way, we’ll meet an emerging discipline that focuses on using machine learning and AI to improve your business: decision intelligence engineering!
Discover our event in València 🇪🇸 thenextweb.com/valencia
Follow us to find out more:
Linkedin - / the-next-web
Instagram - / thenextweb
Facebook - / thenextweb
Twitter - / thenextweb

[@royklaassebos2400]

Three of the key insights: 1. There is a fundamental difference between building and applying Machine Learning algorithms. 2. The use of black-box machine learning models can be justified if you have a proper testing procedure in place. 3. Machine Learning will deliver to you what you ask for, not necessarily what you want. That’s why it’s crucial to be thoughtful about what you ask for.

[@ValiChandiwala]

Danke!

[@jonathanweininger7857]

Cassie is so good at communicating that I’m simultaneously in awe and slightly disappointed that she has already maximized the full potential of how to communicate these concepts.

[@NierAutomata2B]

While being a scientist, she is a masterful presenter with the language skills, body movements and expressions comparable to a stage actress. I mean it in a good way.

[@AbrahamSamma]

This is by far the most down to earth talk on ml 👍

[@dachen4817]

1. Microwave VS kitchen, cook. You don't need first to build a microwave from scratch, then use it to cook. You just need knowing how use it, then do your cook. 2.1 Ingredients -> Appliance -> recipe -> dish 2.2 Data -> Algorithm -> Models -> predictions. 2.3 Building a large scale kitchen is different skill set, compared to build a microwave. 3. ML is a new programming paradigm.Imagine what you want The Island(many smart people just working for you sitting behind screen) to do your repeated work. 3.1 ML research VS ML engineer, Algorithm research VS Applied. 4. you don't understand how your brain work ,how you recognize a cat. But it just work, that's OK. You don't need to do AL from scratch, just dive it , to use it. 5. Testing is vital.

[@devendrasatram5449]

Amazing talk. made me think about the gaps in the ml industry. I first learnt about the p-value significance from Cassie's videos, all her videos are amazing.

[@Prakash-bu3xh]

One of the best talk

[@deepak0417]

One of the best argument and explanation to burst myths behind machine learning and it's real application.

[@kuldeeppilaji]

Such a great talk !!! Thanks Cassie Kozyrkov.

[@gaurangagg]

this is incredible. amazing talk!

[@johnk8174]

"thing-labeling" ... very cool

[@nymthehavoc]

Cassie is doing the same as always: try to convince companies/leaders that the deep level of ML skills are irrelevant for them- as the foundation is already developed by Google- and providing a - somewhat hidden- suggestion that they should use the Google framework which will work out just fine. Due to the relative high level of skillset of the Google staff, I agree with her at some level- but make no mistake- this is an advertisement, which is very well placed and formated. No surprises there, as she is a neuroscientist, who is very skilled predicting what and how you will behave for her promotion.

[@acidtears]

What does her being a neuroscientist have to do with this? Anyone can look up papers on neuroeconomics and how people make decisions... I fail to see how knowing that the amygdala or ventromedial PFC are involved in decision making somehow makes her more skilled at "predicting" people's behaviour.

[@acidtears]

To an extent I agree with you though. It is well worth the effort and the time to study and learn ML. Knowing what is under the hood can be incredibly helpful when resolving issues. Here's to hoping that too many companies will use these pre-packaged frameworks and libraries and will rely on people with deep knowledge to maintain and extent those systems!

[@parvanehshafiei1]

I don't know you get this from her talk, it was an easy explanation of Ml & AI, anyway if u think she is trying to convince u to try Google products i don't see any harm on that!

[@purefatdude2]

Lol, study of human behavior due to external factors is social psychology, not neuroscience. Neuroscience studies the relationship between brain activity and human behavior. Epic fail.

[@kw6143]

@@purefatdude2 actually her PhD is in ‘psychology and neuroscience’

[@SergioArroyoSailing]

Fantastic talk, superbly delivered. Kudos to the speaker. I love how this speech resonates on so many levels

[@diencai1812]

Amazing speaker

[@bhaarathtj6983]

Great talk in a language that's easy to understand

[@sarnathk1946]

On FDA, we trust!

[@kimchiquesadilla]

14:50 fangirling

[@kaidszory8031]

So coollll

[@AnuragTK]

Good talk, but I feel like it is easier for her to put out this perspective of hers when she is at the cutting edge of ML/AI. Everyone does not have the same bird's eye point of view of the domain.

[@pedroavila1833]

Jeez, there's an ad about ML courses on Udemy before this talk...exactly what the course is saying you SHOULDN'T do. Classic.

[@NicO-cm2xo]

Ineffable = unscrewable???

[@miriistina]

I'm surprised it's not Kozyrkova....

[@salsaestonia]

Great talk, but that stage is horrible for speakers and also for the audience.

[@SY-me5rk]

Google doing google things as usual

[@olafpenne]

Bit outdated story. Fortunately nowadays there is much more focus on the transparency of ML models.

[@robertboland6963]

A bit long winded & rambling talk as not all ML perspectives are one and the same. Themes are ping ponged about. Audience looks disengaged.

[@bolso66]

Talks that are supposed to be riveting shows are just sh it

[@DrJohnPollard]

Honestly, I didn't hear one cogent idea out of this chick's rap. Smoke and mirrors.