text
stringlengths
11
441
[2323.72 --> 2326.52] of their Face ID
[2326.52 --> 2329.98] and has raised some fairly interesting
[2329.98 --> 2332.72] points here. Namely,
[2333.22 --> 2336.38] so Apple has stated that all FacePrint data will be stored locally
[2336.38 --> 2339.16] on an individual's device as opposed to being sent to the cloud.
[2340.18 --> 2343.74] So, is it currently possible, either remotely or through physical access
[2343.74 --> 2347.48] to the device, for either Apple or a third party to extract and obtain
[2347.48 --> 2349.88] usable FacePrint data from the iPhone X?
[2350.92 --> 2353.72] Is there any foreseeable reason why Apple would decide to begin
[2353.72 --> 2354.94] storing such data remotely?
[2355.64 --> 2358.10] Apple has stated it used more than one billion images
[2358.10 --> 2359.82] in developing the Face ID algorithm.
[2360.38 --> 2362.80] Where did these one billion face images come from?
[2362.84 --> 2363.76] That's less relevant to me.
[2364.40 --> 2365.80] Really? You don't think that's interesting?
[2367.08 --> 2368.06] It could be,
[2368.44 --> 2370.24] but the answer could be very mundane.
[2370.88 --> 2371.82] It could be shutterstock.
[2372.52 --> 2374.46] It could be just public images of faces.
[2374.46 --> 2376.50] Where would you get a billion images of faces?
[2376.80 --> 2377.34] The internet.
[2377.98 --> 2378.40] Well, yeah.
[2378.40 --> 2379.14] The whole internet.
[2380.04 --> 2383.26] Yeah, but then how would you guarantee that you have the rights
[2383.26 --> 2384.00] to use it?
[2384.60 --> 2386.60] Because the whole terms of service
[2386.60 --> 2389.70] where you can collect non-personally identifying information,
[2389.92 --> 2391.88] the thing that people usually agree to
[2391.88 --> 2394.44] is that it's not supposed to be personally identifying.
[2394.58 --> 2396.36] Your face is by definition
[2396.36 --> 2397.96] personally identifying.
[2397.96 --> 2401.04] What steps did Apple take to ensure its system
[2401.04 --> 2402.88] was trained on a diverse set of faces
[2402.88 --> 2404.36] in terms of race, gender, and age?
[2405.00 --> 2407.90] How are they protecting against racial, gender, or age bias
[2407.90 --> 2408.56] in Face ID?
[2409.26 --> 2411.88] In the unveiling of the iPhone X,
[2411.96 --> 2413.56] Apple made numerous assurances
[2413.56 --> 2415.36] about the accuracy and sophistication.
[2415.74 --> 2417.40] Please describe again all the steps
[2417.40 --> 2418.28] that Apple has taken
[2418.28 --> 2420.62] to ensure that Face ID can distinguish
[2420.62 --> 2422.98] an individual's face from a photograph or mask,
[2423.00 --> 2423.36] for example.
[2423.36 --> 2427.20] Apple, I would love to do a death mask.
[2428.36 --> 2431.96] And then, I don't know exactly how we'd do it.
[2432.04 --> 2433.94] I would love to have a Linus mask.
[2435.60 --> 2438.38] Can I, uh, if we make one for the video, okay?
[2438.64 --> 2440.12] No, you can't have it.
[2440.52 --> 2441.46] You're never going to use it again.
[2441.54 --> 2442.82] That is so weird.
[2443.90 --> 2445.02] Drop all, if we made Linus masks.
[2445.02 --> 2446.28] Would you wear it while you watch?
[2449.58 --> 2451.80] You can't drop F-bombs, you can say that.
[2451.80 --> 2454.02] Okay, if we had one in the merch store.
[2454.08 --> 2455.80] Look, we have to draw the line somewhere.
[2455.92 --> 2457.80] How many people would be Linus for Halloween?
[2458.00 --> 2459.90] We all know I wasn't going to stop saying
[2459.90 --> 2461.76] weird sexual stuff, so.
[2462.62 --> 2464.56] Okay, at least that can go over somebody's head.
[2464.72 --> 2466.18] Whereas F-bombs are just F-bombs.
[2466.52 --> 2467.16] Yeah, see?
[2469.44 --> 2470.42] The answer is yes.
[2470.58 --> 2471.06] Yes, I would.
[2471.18 --> 2472.14] It would enhance the experience.
[2472.58 --> 2474.64] Can Apple confirm that it has no plans
[2474.64 --> 2476.36] to use face print data for any purpose
[2476.36 --> 2477.88] other than the operation of Face ID?
[2478.00 --> 2478.96] These are all good questions.
[2479.14 --> 2479.26] That's a big one.
[2479.26 --> 2481.58] The biggest one for me is,
[2481.66 --> 2483.86] is Face ID really going to be more convenient
[2483.86 --> 2485.82] than Touch ID?
[2486.10 --> 2487.70] And that's one that I'm having some trouble with.
[2488.14 --> 2491.44] Because they did show it operating
[2491.44 --> 2493.58] not directly staring at the phone,
[2493.72 --> 2495.50] which is the biggest issue I have right now
[2495.50 --> 2496.48] with Iris ID.
[2496.98 --> 2498.74] By the time I go like this,
[2498.84 --> 2499.56] and I swear,
[2499.68 --> 2502.00] probably a good 40 to 50% of the time I use it,
[2502.06 --> 2502.26] I'm like...
[2503.54 --> 2504.22] That's gash.
[2504.32 --> 2504.64] You don't want to.
[2504.64 --> 2505.50] Playing around with it.
[2505.50 --> 2507.60] A lot of the time,
[2507.64 --> 2509.84] it would be faster for me to just key in a pen.
[2510.16 --> 2512.46] Because fingerprint unlock on the back of your device,
[2512.66 --> 2514.08] which you iPhone people aren't used to,
[2514.52 --> 2515.66] you can unlock that sucker
[2515.66 --> 2516.88] while it's still in your pocket,
[2516.94 --> 2517.88] and you bring it to your face,
[2517.90 --> 2518.82] it's already there.
[2519.48 --> 2521.64] So you're saying that they demonstrated this thing,
[2521.78 --> 2522.78] and at up two angles,
[2523.40 --> 2524.52] it can already be unlocking
[2524.52 --> 2525.60] before it's even right directly in front of you?
[2525.60 --> 2526.70] Well, they had it sitting on the desk.
[2526.82 --> 2528.22] Like, they were simulating the,
[2528.22 --> 2528.70] you know,