TED2014
Del Harvey: Protecting Twitter users (sometimes from themselves)
Del Harvey: Čudnovatost razmjera u Twitteru
Filmed:
Readability: 4.4
993,199 views
Kada se stotine tisuća Tweetova ispaljuje svake sekunde, šanse jedan naprama milijun -- uključujući nevjerojatno zvučećih scenarija koji bi mogli naškoditi korisnicima --događaju se otprilike 500 puta dnevno. Za Del Harvey, koja vodi Twitterov Tim za Sigurnost i Povjerenje, ove šanse nisu dobre. Voditeljica osiguranja provodi dane razmišljajući kako spriječiti najgore moguće scenarije, istovremeno dajući glas ljudima diljem planete. Sa bezizraznim humorom, daje nam prozor u to kako 240 milijuna korisnika održava sigurnima.
Del Harvey - Security maven
Del Harvey works to define policy and to ensure user safety and security in the challenging realm of modern social media. Full bio
Del Harvey works to define policy and to ensure user safety and security in the challenging realm of modern social media. Full bio
Double-click the English transcript below to play the video.
00:12
My job at Twitter
0
984
1291
Moj posao u Twitteru
00:14
is to ensure user trust,
1
2275
1978
je osigurati povjerenje korisnika,
00:16
protect user rights and keep users safe,
2
4253
2837
zaštiti prava korisnika i očuvati sigurnost korisnika,
00:19
both from each other
3
7090
1260
jedne od drugih
00:20
and, at times, from themselves.
4
8350
3899
i, ponekad, od sebe samih.
00:24
Let's talk about what scale looks like at Twitter.
5
12249
4275
Pričajmo kako razmjer izgleda u Twitteru.
00:28
Back in January 2009,
6
16524
2870
U siječnju 2009.
00:31
we saw more than two million new tweets each day
7
19394
3331
vidjeli smo više od dva milijuna novih tweetova svakog dana
00:34
on the platform.
8
22725
1764
na platformi.
00:36
January 2014, more than 500 million.
9
24489
5908
Siječanj 2014., više od 500 milijuna.
00:42
We were seeing two million tweets
10
30397
2492
Vidimo dva milijuna tweetova
00:44
in less than six minutes.
11
32889
2176
u manje od šest minuta.
00:47
That's a 24,900-percent increase.
12
35065
6984
To je povećanje od 24900 posto.
00:54
Now, the vast majority of activity on Twitter
13
42049
3253
Sad, velika većina aktivnosti na Twitteru
00:57
puts no one in harm's way.
14
45302
1503
nikome ne škodi.
00:58
There's no risk involved.
15
46805
1935
Ne postoji rizik.
01:00
My job is to root out and prevent activity that might.
16
48740
5753
Moj posao je pronaći i spriječiti aktivnost koja bi nekom mogla naškoditi.
01:06
Sounds straightforward, right?
17
54493
1973
Zvuči jednostavno, zar ne?
01:08
You might even think it'd be easy,
18
56466
1152
Možete čak misliti da bi to bilo lako,
01:09
given that I just said the vast majority
19
57618
2170
s obzirom na to da sam rekla da je velika većina
01:11
of activity on Twitter puts no one in harm's way.
20
59788
3810
aktivnosti na Twitteru bezazlena.
01:15
Why spend so much time
21
63598
2169
Zašto potrošiti toliko vremena
01:17
searching for potential calamities
22
65767
2743
tražeći potencijalne nezgode
01:20
in innocuous activities?
23
68510
2900
u neškodljivim aktivnostima?
01:23
Given the scale that Twitter is at,
24
71410
2940
Uzimajući u obzir razmjer na kojem je Twitter,
01:26
a one-in-a-million chance happens
25
74350
2357
situacija jedan u milijun događa se
01:28
500 times a day.
26
76707
4876
500 puta na dan.
01:33
It's the same for other companies
27
81583
1445
To je točno i za druge tvrtke
01:35
dealing at this sort of scale.
28
83028
1471
koje rade na ovom razmjeru.
01:36
For us, edge cases,
29
84499
1708
Za nas, rubni slučajevi,
01:38
those rare situations that are unlikely to occur,
30
86207
3625
one rijetke situacije koje su neizgledne da se dogode,
01:41
are more like norms.
31
89832
2622
su više kao norma.
01:44
Say 99.999 percent of tweets
32
92454
3942
Recimo da 99,999 posto tweetova
01:48
pose no risk to anyone.
33
96396
1888
ne predstavljaju prijetnju nikome.
01:50
There's no threat involved.
34
98284
1066
Ne postoji nikakva prijetnja.
01:51
Maybe people are documenting travel landmarks
35
99350
2954
Možda ljudi bilježe putne znamenitosti
01:54
like Australia's Heart Reef,
36
102304
1963
Kao što je Australski Heart Reef
01:56
or tweeting about a concert they're attending,
37
104267
2921
ili tweetaju o koncertu na kojem su,
01:59
or sharing pictures of cute baby animals.
38
107188
4747
ili dijele slike slatkih životinjskih mladunaca.
02:03
After you take out that 99.999 percent,
39
111935
4509
Kada oduzmete tih 99,999 posto,
02:08
that tiny percentage of tweets remaining
40
116444
3529
taj mali postotak tweetova koji ostaje
02:11
works out to roughly
41
119973
2389
daje otprilike
02:14
150,000 per month.
42
122362
3475
150,000 mjesečno.
02:17
The sheer scale of what we're dealing with
43
125837
2456
Čisti razmjer onoga čime se bavimo
02:20
makes for a challenge.
44
128293
2312
predstavlja izazov.
02:22
You know what else makes my role
45
130605
1178
Znate što još čini moju ulogu
02:23
particularly challenging?
46
131783
3107
posebno izazovnom?
02:26
People do weird things.
47
134890
5123
Ljudi rade čudne stvari.
02:32
(Laughter)
48
140013
1829
(Smijeh)
02:33
And I have to figure out what they're doing,
49
141842
2391
I moram otkriti što oni to rade,
02:36
why, and whether or not there's risk involved,
50
144233
2249
zašto, i postoji li ili ne rizik,
02:38
often without much in terms of context
51
146482
2168
često bez puno konteksta
02:40
or background.
52
148650
1847
ili pozadine.
02:42
I'm going to show you some examples
53
150497
2077
Pokazat ću vam neke primjere
02:44
that I've run into during my time at Twitter --
54
152574
2005
na koje sam naišla za vrijeme rada u Twitteru --
02:46
these are all real examples —
55
154579
1620
ovo su sve stvarni primjeri --
02:48
of situations that at first seemed cut and dried,
56
156199
2653
situacija koje su se činile jednostavne na početku
02:50
but the truth of the matter was something
57
158852
1643
ali stvarna priroda stvari bila je nešto
02:52
altogether different.
58
160495
1550
potpuno drugačije,
02:54
The details have been changed
59
162045
1977
Detalji su promijenjeni
02:56
to protect the innocent
60
164022
1257
kako bi zaštitili nevine
02:57
and sometimes the guilty.
61
165279
3233
i ponekad krivce.
03:00
We'll start off easy.
62
168512
3005
Počet ćemo lagano.
03:03
["Yo bitch"]
63
171517
1793
["Hej kujo"]
03:05
If you saw a Tweet that only said this,
64
173310
3228
Kada biste vidjeli Tweet koji kaže samo ovo,
03:08
you might think to yourself,
65
176538
1694
pomislili biste,
03:10
"That looks like abuse."
66
178232
1653
"Ovo izgleda kao zlostavljanje."
03:11
After all, why would you
want to receive the message,
want to receive the message,
67
179885
3107
Na kraju krajeva, zašto biste htjeli primiti poruku.
03:14
"Yo, bitch."
68
182992
2218
"Hej, kujo."
03:17
Now, I try to stay relatively hip
69
185210
4663
Sad, pokušavam ostati relativno u tijeku
03:21
to the latest trends and memes,
70
189873
2512
sa najnovijim trendovima i memeima,
03:24
so I knew that "yo, bitch"
71
192385
2704
pa sam znala da "Hej, kujo"
03:27
was also often a common greeting between friends,
72
195089
3154
je također učestali pozdrav među prijateljima,
03:30
as well as being a popular "Breaking Bad" reference.
73
198243
4262
kao i popularna referenca na "Breaking Bad".
03:34
I will admit that I did not expect
74
202505
2487
Priznat ću da nisam očekivala
03:36
to encounter a fourth use case.
75
204992
2841
susresti četvrti slučaj korištenja.
03:39
It turns out it is also used on Twitter
76
207833
3104
Ispada da se također koristi na Twitteru
03:42
when people are role-playing as dogs.
77
210937
3062
kada ljudi glume pse.
03:45
(Laughter)
78
213999
5279
(Smijeh)
03:51
And in fact, in that case,
79
219278
1666
I zapravo, u tom slučaju,
03:52
it's not only not abusive,
80
220944
1609
ne samo da nije zlostavljanje
03:54
it's technically just an accurate greeting.
81
222553
3139
nego je to tehnički ispravan pozdrav.
03:57
(Laughter)
82
225692
2889
(Smijeh)
04:00
So okay, determining whether or not
83
228581
2071
Dakle, utvrđivanje je li ili nije
04:02
something is abusive without context,
84
230652
1848
nešto uvredljivo bez konteksta,
04:04
definitely hard.
85
232500
1592
definitivno teško.
04:06
Let's look at spam.
86
234092
2717
Pogledajmo spam.
04:08
Here's an example of an account engaged
87
236809
1960
Ovo je primjer računa koji je uključen
04:10
in classic spammer behavior,
88
238769
1668
u klasično spammersko ponašanje,
04:12
sending the exact same message
89
240437
1559
šaljući potpuno jednaku poruku
04:13
to thousands of people.
90
241996
1804
tisućama ljudi.
04:15
While this is a mockup I put
together using my account,
together using my account,
91
243800
2793
Iako je ovo samo primjer koji sam napravila koristeći svoj račun,
04:18
we see accounts doing this all the time.
92
246593
3001
vidimo račune kako to rade cijelo vrijeme.
04:21
Seems pretty straightforward.
93
249594
1979
Čini se poprilično jednostavno.
04:23
We should just automatically suspend accounts
94
251573
2053
Jednostavno bi trebali suspendirati račune
04:25
engaging in this kind of behavior.
95
253626
3307
koji se bave ovakvim stvarima.
04:28
Turns out there's some exceptions to that rule.
96
256933
3210
Ispada da postoje neke iznimke tome pravilu.
04:32
Turns out that that message
could also be a notification
could also be a notification
97
260143
2883
Ispada da poruke mogu biti i obavijest
04:35
you signed up for that the International
Space Station is passing overhead
Space Station is passing overhead
98
263026
3889
za koju ste se prijavili kako bi znali
da Međunarodna Svemirska Postaja prolazi
da Međunarodna Svemirska Postaja prolazi
04:38
because you wanted to go outside
99
266915
1846
jer ste htjeli izaći van
04:40
and see if you could see it.
100
268761
1948
i pokušati vidjeti je.
04:42
You're not going to get that chance
101
270709
1225
Nećete dobiti tu priliku
04:43
if we mistakenly suspend the account
102
271934
1847
ako mi greškom suspendiramo račun
04:45
thinking it's spam.
103
273781
2266
misleći da je to spam.
04:48
Okay. Let's make the stakes higher.
104
276047
3526
Ok. Povisimo uloge.
04:51
Back to my account,
105
279573
1916
Natrag na moj račun,
04:53
again exhibiting classic behavior.
106
281489
3505
ponovno prikazuje klasično ponašanje.
04:56
This time it's sending the same message and link.
107
284994
2643
Ovoga puta šalje istu poruku i poveznicu.
04:59
This is often indicative of
something called phishing,
something called phishing,
108
287637
2774
Ovo je često indikator nečega što se zove "phishing"
05:02
somebody trying to steal another
person's account information
person's account information
109
290411
3178
odnosno kada netko pokušava ukrasti
informacije o računu druge osobe
informacije o računu druge osobe
05:05
by directing them to another website.
110
293589
2203
usmjeravajući ih na drugu web stranicu.
05:07
That's pretty clearly not a good thing.
111
295792
4194
To očito nije dobra stvar.
05:11
We want to, and do, suspend accounts
112
299986
1930
Mi želimo, i to i činimo, suspendirati račune
05:13
engaging in that kind of behavior.
113
301916
2624
koji se bave ovakvim stvarima.
05:16
So why are the stakes higher for this?
114
304540
3247
Zašto su onda ulozi veći u ovakvim situacijama?
05:19
Well, this could also be a bystander at a rally
115
307787
2999
Pa, ovo bi mogao biti promatrač na skupu
05:22
who managed to record a video
116
310786
1910
koji je uspio snimiti video
05:24
of a police officer beating a non-violent protester
117
312696
3270
policajca koji tuče prosvjednika
05:27
who's trying to let the world know what's happening.
118
315966
2975
koji pokušava svijetu reći što se događa.
05:30
We don't want to gamble
119
318941
1643
Ne želimo se kockati
05:32
on potentially silencing that crucial speech
120
320584
2517
s tim da nehotično utišamo taj krucijalni govor
05:35
by classifying it as spam and suspending it.
121
323101
2929
klasificirajući ga kao spam i suspendirajući ga.
05:38
That means we evaluate hundreds of parameters
122
326030
2879
To znači da procjenjujemo stotine parametara
05:40
when looking at account behaviors,
123
328909
1688
kada gledamo ponašanja računa,
05:42
and even then, we can still get it wrong
124
330597
2016
i čak i tada možemo pogriješiti
05:44
and have to reevaluate.
125
332613
2236
i moramo ponovno procjenjivati.
05:46
Now, given the sorts of challenges I'm up against,
126
334849
3708
Sad, uzimajući u obzir izazove s kojima se suočavam,
05:50
it's crucial that I not only predict
127
338557
2696
ključno je ne da samo predvidim
05:53
but also design protections for the unexpected.
128
341253
3784
nego i dizajniram zaštitu od neočekivanog.
05:57
And that's not just an issue for me,
129
345037
2342
I to nije samo problem za mene,
05:59
or for Twitter, it's an issue for you.
130
347379
2087
ili Twitter, to je vaš problem.
06:01
It's an issue for anybody who's building or creating
131
349466
2406
To je problem za svakoga tko gradi ili stvara
06:03
something that you think is going to be amazing
132
351872
1925
nešto za što mislite da će biti odlično
06:05
and will let people do awesome things.
133
353797
2789
i dopustit će ljudima da čine odlične stvari.
06:08
So what do I do?
134
356586
2866
Dakle što ja činim?
06:11
I pause and I think,
135
359452
3318
Zastanem i razmišljam,
06:14
how could all of this
136
362770
2095
kako bi sve ovo moglo
06:16
go horribly wrong?
137
364865
3793
poći strašno po krivu?
06:20
I visualize catastrophe.
138
368658
4453
Vizualiziram katastrofu.
06:25
And that's hard. There's a sort of
139
373111
2463
I to je teško. Postoji određena
06:27
inherent cognitive dissonance in doing that,
140
375574
2848
svojstvena kognitivna disonanca prilikom činjenja toga,
06:30
like when you're writing your wedding vows
141
378422
1812
kao kad pišete svoje svadbene zavjete
06:32
at the same time as your prenuptial agreement.
142
380234
2646
u isto vrijeme kada pišete predbračni ugovor.
06:34
(Laughter)
143
382880
1696
(Smijeh)
06:36
But you still have to do it,
144
384576
2373
Ali to i dalje morate činiti,
06:38
particularly if you're marrying
500 million tweets per day.
500 million tweets per day.
145
386949
4446
osobito ako se udajete za 500 milijuna tweetova po danu.
06:43
What do I mean by "visualize catastrophe?"
146
391395
3097
Što mislim kada kažem "vizualiziranje katastrofe?"
06:46
I try to think of how something as
147
394492
2762
Pokušavam misliti kako nešto
06:49
benign and innocuous as a picture of a cat
148
397254
3228
bezazleno i neškodljivo kao što je slika mačke
06:52
could lead to death,
149
400482
1104
može voditi do smrti,
06:53
and what to do to prevent that.
150
401586
2326
i što učiniti kako bismo to spriječili.
06:55
Which happens to be my next example.
151
403912
2383
Što je moj sljedeći primjer.
06:58
This is my cat, Eli.
152
406295
3110
Ovo je moj mačak, Eli.
07:01
We wanted to give users the ability
153
409405
1981
Htjeli smo dati korisnicima mogućnost
07:03
to add photos to their tweets.
154
411386
2073
da dodaju slike svojim tweetovima.
07:05
A picture is worth a thousand words.
155
413459
1597
Slika vrijedi tisuću riječi.
07:07
You only get 140 characters.
156
415056
2009
A imate samo 140 znakova.
07:09
You add a photo to your tweet,
157
417065
1200
Kada dodate sliku svom tweetu,
07:10
look at how much more content you've got now.
158
418265
3038
pogledajte koliko više sadržaja sada imate.
07:13
There's all sorts of great things you can do
159
421303
1677
Postoji jako puno super stvari koje možete napraviti
07:14
by adding a photo to a tweet.
160
422980
2007
dodajući sliku na tweet.
07:16
My job isn't to think of those.
161
424987
2280
Moj posao nije misliti na njih,
07:19
It's to think of what could go wrong.
162
427267
2747
već misliti na to što bi moglo poći po krivu.
07:22
How could this picture
163
430014
1892
Kako bi ova slika
07:23
lead to my death?
164
431906
3539
mogla dovesti do moje smrti?
07:27
Well, here's one possibility.
165
435445
3160
Pa, evo jedne mogućnosti.
07:30
There's more in that picture than just a cat.
166
438605
3086
Na ovoj slici nije samo moja mačka.
07:33
There's geodata.
167
441691
2092
Postoje geopodatci.
07:35
When you take a picture with your smartphone
168
443783
2212
Kada uslikate fotografiju vašim pametnim telefonom
07:37
or digital camera,
169
445995
1299
ili digitalnom kamerom,
07:39
there's a lot of additional information
170
447294
1654
postoji mnogo dodatnih informacija
07:40
saved along in that image.
171
448948
1616
koje su spremljene u toj slici.
07:42
In fact, this image also contains
172
450564
1932
Zapravo, ova slika također sadrži
07:44
the equivalent of this,
173
452496
1805
ekvivalent ovoga,
07:46
more specifically, this.
174
454301
3079
točnije ovoga.
07:49
Sure, it's not likely that someone's going to try
175
457380
1956
Naravno, malo je vjerojatno da će netko pokušati
07:51
to track me down and do me harm
176
459336
2285
pronaći me i učiniti mi nešto nažao
07:53
based upon image data associated
177
461621
1784
na osnovi podataka sa slike povezane
07:55
with a picture I took of my cat,
178
463405
1948
sa fotografijom moje mačke,
07:57
but I start by assuming the worst will happen.
179
465353
3651
ali počnem pretpostavljati da će se dogoditi najgore.
08:01
That's why, when we launched photos on Twitter,
180
469004
2338
Zato smo, kad smo lansirali fotografije na Twitteru,
08:03
we made the decision to strip that geodata out.
181
471342
3821
odlučili ukloniti sve geopodatke.
08:07
(Applause)
182
475163
5847
(Pljesak)
08:13
If I start by assuming the worst
183
481010
2613
Ako počnem s najgorom pretpostavkom
08:15
and work backwards,
184
483623
947
i radim unatrag,
08:16
I can make sure that the protections we build
185
484570
2553
mogu biti sigurna da će zaštite koje smo izgradili
08:19
work for both expected
186
487123
1768
raditi za očekivane
08:20
and unexpected use cases.
187
488891
2078
i neočekivane slučajeve korištenja.
08:22
Given that I spend my days and nights
188
490969
2945
Uzimajući u obzir da dane i noći provodim
08:25
imagining the worst that could happen,
189
493914
2541
zamišljajući što se najgore može dogoditi
08:28
it wouldn't be surprising if
my worldview was gloomy.
my worldview was gloomy.
190
496455
4257
ne bi bilo čudno da mi je pogled na svijet tmuran.
08:32
(Laughter)
191
500712
1783
(Smijeh)
08:34
It's not.
192
502495
1417
Nije.
08:35
The vast majority of interactions I see --
193
503912
3876
Velika većina interakcija koje vidim --
08:39
and I see a lot, believe me -- are positive,
194
507788
3901
i vidim ih puno vjerujte -- su pozitivne,
08:43
people reaching out to help
195
511689
1924
ljudi pokušavaju pomoći
08:45
or to connect or share information with each other.
196
513613
3448
ili povezati se ili dijeliti informacije jedni s drugima.
08:49
It's just that for those of us dealing with scale,
197
517061
3323
Ali stvar je u tome da mi koji se nosimo s razmjerom,
08:52
for those of us tasked with keeping people safe,
198
520384
3800
mi koji se brinemo da su ljudi sigurni,
08:56
we have to assume the worst will happen,
199
524184
2546
moramo pretpostaviti da će se najgore dogoditi,
08:58
because for us, a one-in-a-million chance
200
526730
4227
jer za nas, šanse jedan naprama milijun
09:02
is pretty good odds.
201
530957
2749
su poprilično dobri izgledi.
09:05
Thank you.
202
533706
1864
Hvala vam.
09:07
(Applause)
203
535570
4000
(Pljesak)
ABOUT THE SPEAKER
Del Harvey - Security mavenDel Harvey works to define policy and to ensure user safety and security in the challenging realm of modern social media.
Why you should listen
As Senior Director of Trust and Safety at Twitter, Del Harvey works to define policy and to ensure user safety and security in the challenging realm of modern social media. Prior to joining Twitter, she spent five years as the co-administrator and law enforcement liaison for a 501(c)3 non-profit charity, working with agencies ranging from local police departments to the FBI, U.S. Marshals and the Secret Service.
Del Harvey | Speaker | TED.com