How to use data to make a hit TV show | Sebastian Wernicke

Does collecting more data lead to better decision-making? Competitive, data-savvy companies like Amazon, Google and Netflix have learned that data analysis alone doesn't always produce optimum results. In this talk, data scientist Sebastian Wernicke breaks down what goes wrong when we make decisions based purely on data - and suggests a brainier way to use it.
TEDTalks is a daily video podcast of the best talks and performances from the TED Conference, where the world's leading thinkers and doers give the talk of their lives in 18 minutes (or less). Look for talks on Technology, Entertainment and Design -- plus science, business, global issues, the arts and much more.
Find closed captions and translated subtitles in many languages at www.ted.com/tra...
Follow TED news on Twitter: / tednews
Like TED on Facebook: / ted
Subscribe to our channel: / tedtalksdirector

[@LaMugre]

sounds like they worked perfectly. Amazon took information from everyone, without selecting a main demographic, and made the most average show possible. Netflix took specific information from specific demographics and made a really good show that apeals to said demographic. The key here is what information to regard.

[@quAdxify]

+La Mugre Exactly, thats the thing that came to my mind as well. I don't know from whom I heard it but there was that Harvard Guy that tried to find the optimal tomato sauce. The final answer was that there is no single global optimum because there are different groups of people that have different tastes, instead he could show that there are multiple optima depending on the respective subgroup. I think Amazon tried to find exactly that global optimum that does not exist. Additionally I don't really think what he concludes is true at all. Our world is stochastic, decision making based on data is stochastic as well. Similarly to human decisions, data decisions are purely stochastic and thus a final verdict would need a really big sample size to grasp the real power of the respective methods. Additionally, he assumes a deterministic world, where given all observations/ data possible we could always predict the outcome, however, quantum physics shows that this is probably just not the case, hence it might be IMPOSSIBLE to make the right decision every time. All we can do is compare human decision making to machines and switch when machines get better (this is already the case in some scenarios, e.g. ANNs showed superhuman performance in traffic sign recognition). I think what he basically tried to say is that machines can't - yet - replace human creativity, which is no surprise but a machine can be really good at making a certain decisions that previously required experts - which he just disregards.

[@zeynand4039]

@@quAdxify it might be impossible for humans to make the "right" decision all the time but human behaviour is highly structured. So maybe looking at the angle of behaviour, you can extract info to align with machines. For example chaos in a mall and how protection signals can be send out, like water supply within the mall can be increased and police being noticed automatically. Things like that can assist humans in making the right decision.

[@Buckets41369]

TED Talks used to blow me away but now they're just a bit average, like a 7.4 on IMDB.

[@WillHirschUK]

+Robbie Lockie It has to be said that TED deserve every bit of discreditation from that confusion though, for allowing it to become such a derivative franchise and then posting x talks on their main TED talk, even putting TED in the title.

[@srimansrini]

In this talk, Data Expert Sebastian Wernicke gives wonderful info about how to wisely use the data to make right decisions. Sometimes, more than the data, our brain gives better options. That is the age-old wisdom. A very useful talk.

[@corpusc]

every Ted talk i've watched recently is 12-16 minutes of jibber jabber that's easily summarized by a tiny common sense idea that everyone is familiar with.

[@andy4an]

+Corpus Callosum yeah, watching the new ones won't get you many great hits. It is better to find the popular ones to get good TED experiences.

[@Chr0nalis]

+Corpus Callosum I disagree.

[@calvinkrist5672]

You say these are ideas everyone is familiar with, and that may be true for many of them- however, there are many I have watched that are not like that. This one specifically points out many places that are not familiar with the concept discussed, with sometimes disturbing results.

[@corpusc]

+Calvin Krist wow, you guys actually have different experiences than me. i am absolutely shocked by that.

[@andy4an]

Calvin Krist did my response sound different? i was trying to agree.

[@jasonm3582]

I see elements of PsychoPass here from 6:30 - 6:52

[@harrison6082]

If Google was able to predict the outbreaks of influenza, what have they got on covid19?

[@TheAdnanSakib]

When he said that's one thing it helps us put it together: the brain. I was just like, wow, I wasted 12 minutes of my life. TED is amazing at misleading titles.

[@brianmyers13]

How did they nail House of Cards and then produce The Ridiculous 6? Hahaha. Strange.

[@EddyMac1903]

Adam Sandler

[@dunnyt1680]

+Future Ruler of Ukraine It's the most watched movie on Netflix already, that's a pretty big success.

[@seanfk]

+Brian Myers you may not have liked the ridiculous 6 but it did really well for netflix

[@MonkeyspankO]

+Brian Myers because House of Cards is not a unique, creative IP. its a re-make of a superior (imho) British show.

[@Maestroxxx1]

+Brian Myers House of Cards was only one success. You would need to consistently use a method successfully multiple times to know whether the success was genuine or a complete fluke. I'm guessing the method used to get HoCs was a complete fluke.

[@jakeisrael7899]

Pretty informative would watch again

[@Overonator]

Data without theories is meaningless. You should talk to philosophers of science who have known this for decades now. But it appears that some are only know realizing it.

[@mcdzmcdz6342]

+Overonator so is the data in a large bank account meaningless without a theory?

[@Overonator]

I assume you mean the numbers in a bank account. And yes they are meaningless without Number Theory.

[@mcdzmcdz6342]

Well your assumptions hold no valid supporting evidence.

[@Overonator]

McDz McDz Can you state my assumptions that hold no valid supporting evidence?

[@Overonator]

***** That's a cool idea. Philosophers need to poke their head out of academia and engage with the public. However what incentive do they have for doing so? That could be a reason why so few actually do this.

[@dannyrowleyyt]

A TED Talk in stating the bloody obvious!

[@joshwillis1726]

I wonder then about how effective the current path of AI will be in the future. Right now we basically just teach a computer using thousands or millions of examples and then use that training and some clever algorithms to make a prediction. Works pretty well for solving problems like vision or engineering, but what about communication? That problem is way more complicated than simple analysis of statistics.

[@Queerbbby]

very good talk!

[@barkspawn]

I'd rather spend five years in prison than watch the Dexter finale again.

[@4amcuriosity162]

Very smart man. Thankful for this video .

[@adrienperie6119]

And now we know why TV has been sucking so badly

[@success_chemistry]

Perhaps risk can also be seen as controversy

[@hoorooblu]

Netflix chose to make House of Cards as a remake of a British 1990 version. Netflix was not making data-only decision. It could see a successful prototype.

[@tobiojo10]

(frank underwood was a congressman not a senator *) otherwise good TED talk

[@LaMugre]

missleading tittle is missleading

[@rickloyd8208]

Based on my experience, data analyses never helped. I feel like it was created to impress stupid boss or detect problem (if something do not fits general stats)

[@philspencer8504]

good talk

[@ZioStalin]

Arrow demonstrated that no election system can consistently represent the preferences of a population. This applies to economics, to data analysis (which is comparable to an election system) and of course this is the second best argument to prove the fallacy of democracy. What is the first best argument against democracy? Easy: a 15' talk with the average elector.

[@willlundgren6440]

this is fire

[@BenJordy]

Love this

[@opl500]

"engineering success"

[@lafcadiothelion]

House of Cards is a remake of a British show with the same title... so Netflix just saw that and then remade it...

[@quizzer]

Misleading title. There's virtually no content here.

[@andy-for-real]

A speech with very little substance. not much research at all, but 3 examples. 12 mins of talking about data without data to back things up. only concluded to use an expert's brain after doing your analysis with data. Well, doesn't seem inspiring or groundbreaking to me at all.

[@Taric25]

What an unbelievably misleading title

[@muddi900]

No House of Cards was awful from the start

[@thegreat5250]

Hey! Someone, who are better than me in English! Please, take me a chance to understand the idea of this video and... Who is Roy?

[@Chr0nalis]

+The Great Gatsby Try using subtitles in your language. I think youtube is now capable of that.

[@thegreat5250]

Shagas Heizenberg I was trying to translate English subtitles, but now I can't do this at the enough speed. So, I know about auto translate, but when KZbin trying to translate English to Russian.. It looks SO terrible...

[@Chr0nalis]

Тагда англиский повтори )

[@user-px2ov7gi7l]

+john smith Can you explain what Russia did to Crimea?

[@user-px2ov7gi7l]

john smith How did Russia steal Crimea?

[@Nootathotep]

Data isn't enough

[@LeonidasGGG]

It's "Appleseed" all over again.

[@user-wh5ir6xg4v]

Do you like sushi?

[@TheStickCoverBand]

8 is average? :o

[@vrishabhlakhani]

this guy looks like Edward snowden

[@booksRbad]

about three good sentences in this whole talk. video could have been cut down to about 40 seconds.

[@coolmdj111]

Wow!... Guys, I didn't know what this video would be about but I happened to post a video on House of Cards like a month back. Even if you don't follow the series yet, perhaps check it out? I'm sure you won't be disappointed... I'd appreciate your feedback. *#FU2016*

[@IDislikeTheNewYoutube]

What was the overall point of this video? I think its a muddled at best idea.

[@ShahFaisalz]

v

[@killthewall101]

meh

[@JinFrankster]

This is y Tedtalk is losing viewers.

[@valaks36]

///////////////////////////////////////////////// this is my fight song/