Latest / Connecting the Dots / AI Hits the Big Screen, Earthquake Alert Gaps, and Tariffs Reshape Fast Fashion

Transcript

AI Hits the Big Screen, Earthquake Alert Gaps, and Tariffs Reshape Fast Fashion · Connecting the Dots

• 0:00
  Today's tech news moves fast, but understanding
• 0:03
  it shouldn't be hard. That's why Snarfl Solutions
• 0:06
  Group brings you Connecting the Docs, where we
• 0:09
  break down what's happening in tech and why it
• 0:11
  matters to you. Now, let's dive in with your
• 0:14
  hosts. Welcome, everyone. Thanks for tuning in.
• 0:17
  We're here to connect the docs on today's tech
• 0:19
  news using the latest insights from our Morningstar
• 0:22
  report. We'll break down what's happening and
• 0:25
  really get into why it matters for you. It is
• 0:27
  Monday, July 28th. 2025 it is indeed and it's
• 0:31
  a lovely sunny day out here in Sacramento Going
• 0:34
  up to about 92 Fahrenheit later. Pretty nice.
• 0:36
  Sounds perfect. Okay quick look at the markets
• 0:39
  this morning The Dow is up just a tiny bit plus
• 0:42
  15 points or so S &P 500 also slightly up about
• 0:47
  point one three percent Bitcoin though is down
• 0:49
  over one percent sitting around a hundred and
• 0:52
  eighteen thousand. Okay bit of a dip there Yeah,
• 0:55
  and closer to home for our baseball fans Tough
• 0:59
  night for the Giants. Oh, no, don't tell me.
• 1:01
  Yeah, lost 3 -5 to the Mets yesterday, and the
• 1:03
  Mets completed the sweep, so rough weekend for
• 1:06
  the Giants faithful. That's a tough break. Always
• 1:09
  hard to see a sweep. Definitely. But all right,
• 1:11
  let's ship gears. We've got three big topics
• 1:13
  from the Morningstar report we want to untag
• 1:14
  today. Right. First, we're looking at AI and
• 1:17
  filmmaking, some really interesting developments
• 1:19
  there. Then a pretty serious update. on Google's
• 1:23
  earthquake alert system. Yeah, that was crucial.
• 1:25
  And finally, we'll talk about how new tariffs
• 1:27
  are hitting fast fashion, specifically companies
• 1:30
  like Sheen. OK, let's jump into that first one.
• 1:33
  AI in filmmaking. I saw this news about IMAX
• 1:36
  teaming up with Runway AI. They're screening
• 1:38
  AI Film Festival finalists. That's right. Ten
• 1:41
  finalists at ten IMAX locations across the U
• 1:44
  .S. They're a pretty big signal showing how fast
• 1:46
  this tech is moving into, well, more mainstream
• 1:49
  spaces. It feels like it exploded, right? Runway's
• 1:52
  festival is only, what, three years old? Exactly,
• 1:54
  third year. And the growth is just wow. In 2023,
• 1:58
  they had around 300 films submitted. OK. This
• 2:00
  year, six thousand. Whoa. Six thousand that's
• 2:04
  huge. It's massive the CEO Cristobal Valenzuela
• 2:07
  He put it really well said something like three
• 2:10
  years ago. This was such a crazy idea today Millions
• 2:13
  of people are making billions of videos using
• 2:15
  tools. We only dreamed of incredible leap So
• 2:18
  what are these AI films actually like is it all
• 2:21
  just you know? Computers generating movies start
• 2:24
  to finish. That's a good question. It's actually
• 2:26
  more nuanced the films are really varied the
• 2:29
  top winner total pixel space is like almost nine
• 2:32
  and a half minutes long. Okay. And it explores
• 2:35
  this idea of how many images can even exist digitally,
• 2:38
  uses math for these huge numbers, described as
• 2:41
  having stunning visuals, kind of reality bending
• 2:44
  stuff. Huh, interesting. Then you've got totally
• 2:46
  different things like Jailbird, which is like
• 2:48
  a documentary about a chicken. in a UK prison,
• 2:52
  from the chicken's point of view. Okay, wow.
• 2:54
  A chicken documentary. Yeah. And another called
• 2:56
  One About Space Travel. The key thing is they
• 2:58
  don't have to be 100 % AI generated. Yeah. Mixed
• 3:01
  media is actually encouraged. So blending real
• 3:04
  shots with AI stuff, it's more about experimenting.
• 3:07
  Gotcha. So it's a tool, not necessarily a replacement
• 3:09
  for the filmmaker. How's the big industry like
• 3:12
  Hollywood taking all this? Are they seeing it
• 3:13
  as, you know, a threat or more of an opportunity?
• 3:16
  Well... From the executive side, it seems mostly
• 3:19
  like an opportunity. You know, efficiency, innovation.
• 3:24
  Professor Joshua Glick mentioned that AI is already
• 3:26
  way more involved than people think. Right. Not
• 3:29
  just the flashy stuff. Exactly. Not just de -aging
• 3:32
  actors or big CGI. It's in post -production,
• 3:35
  editing, digital touch -ups, even just sorting
• 3:38
  through mountains of footage. Studios see it
• 3:41
  as a way to speed things up, do things in minutes
• 3:43
  that took hours. Okay, efficiency makes sense
• 3:45
  from a business perspective, but that obviously
• 3:48
  brings up the other side, the concerns from workers,
• 3:51
  right? Yeah. Unions must have things to say.
• 3:52
  Oh, absolutely. Unions like IATA and SAG -AFTRA,
• 3:56
  they're very clear. They're okay with tech that
• 3:58
  helps tell stories, but they're pushing back
• 4:00
  hard against anything that undermines jobs or
• 4:02
  rights. They want guardrails. Makes sense, need
• 4:05
  protection. Yeah. But then you have the counterargument,
• 4:07
  like from Runway CEO Valenzuela, saying AI can
• 4:10
  help creatives, he kind of compares the fear
• 4:13
  now to fears people had about earlier film tech
• 4:16
  breakthroughs. History repeating itself, maybe.
• 4:19
  Maybe. And filmmaker Flying Lotus, Stephen Ellison,
• 4:22
  he talked about this idea of play around and
• 4:25
  find out. Like AI lets you stumble onto cool
• 4:28
  stuff, these hallucinatory dreamlike results
• 4:31
  you didn't plan. So it's like collaborating with
• 4:34
  the randomness of it. Kind of, yeah. It's less
• 4:36
  about total control and more about seeing what
• 4:39
  emerges. It's this interesting mix of human intent
• 4:42
  and well. AI weirdness that is where it gets
• 4:46
  really interesting that blend but you know thinking
• 4:49
  about AI relying on it for creative stuff is
• 4:51
  one thing what happens when we rely on it for
• 4:53
  something critical like safety and it well it
• 4:56
  drops the ball that's a perfect transition actually
• 4:58
  because our next story is exactly about that
• 5:01
  Google's Android earthquake alert system the
• 5:04
  AEA right the situation in Turkey last year they've
• 5:07
  admitted some pretty serious failures haven't
• 5:09
  they they have a huge failure really the system
• 5:11
  should have sent out about 10 million take action
• 5:14
  alerts the highest level warning during those
• 5:16
  devastating earthquakes 10 million and how many
• 5:19
  actually went out 469 wait 469 instead of 10
• 5:24
  million yeah instead it sent about half a million
• 5:27
  of the lower level be aware notifications wow
• 5:30
  that's That's staggering, especially given the
• 5:33
  timing, 4 .17 a .m., and the terrible human cost
• 5:37
  over 55 ,000 lives lost. Those high -level alerts
• 5:40
  could have made such a difference. Absolutely
• 5:43
  critical. Just to clarify the difference, take
• 5:45
  -action alerts are designed to be impossible
• 5:47
  to miss. Loud sound, overrides do not disturb,
• 5:50
  takes over your screen. Vital for waking people
• 5:53
  up. Be aware is much less intrusive. for lighter
• 5:55
  shaking. The AEA system itself, launched in 2020,
• 5:59
  is clever. It uses the motion sensors, the accelerometers,
• 6:02
  and Android phones. Right, crowdsourcing the
• 6:03
  detection. Exactly. It's active in almost 100
• 6:05
  countries. Yeah. But in this case, that massive
• 6:08
  network just didn't deliver the right warning
• 6:10
  to the vast majority of people who needed it
• 6:12
  most. So what went wrong? Why did it fail so
• 6:14
  badly? Google initially said it performed well,
• 6:16
  didn't they? They did initially, which makes
• 6:19
  the later admission even more stark. In a Science
• 6:22
  Magazine article, they finally detailed the limitations
• 6:24
  to the detection algorithms. Essentially, the
• 6:28
  system drastically underestimated the magnitude
• 6:31
  of that first 7 .8 quake. How badly? It judged
• 6:34
  it as being somewhere between 4 .5 and 4 .9 magnitude.
• 6:38
  Oh, wow. That's a huge difference. Huge. And
• 6:41
  even the second major quake later that day was
• 6:44
  underestimated. Though slightly more alerts went
• 6:46
  out than about 8 ,000 take action ones, still
• 6:49
  far short of what was needed. And it took them
• 6:50
  two years to release these details. Two years.
• 6:53
  Which is drawn criticism, as you can imagine.
• 6:56
  Yeah, understandably. So have they fixed it?
• 6:58
  What happens now? Well, they say they've revised
• 7:00
  the algorithms, they ran simulations with the
• 7:02
  new model, and now it shows it would have set
• 7:04
  the 10 million take action alerts and even more
• 7:06
  be -aware ones. Okay, so it should work better
• 7:09
  now. It should. But Google also points out that
• 7:12
  tuning these systems for massive quakes is a
• 7:14
  challenge for everyone in the field. It really
• 7:17
  highlights this black box issue with critical
• 7:19
  AI, doesn't it? How so? Well, even the developers
• 7:22
  sometimes struggle to pinpoint exactly why a
• 7:24
  failure occurred in such complex systems. And
• 7:27
  when lives are at stake, that's a serious problem.
• 7:30
  It raises questions about trust regulation and
• 7:34
  whether we're sometimes placing too much faith
• 7:36
  in tech that's still evolving. Yeah. And Google
• 7:38
  themselves say it's not supposed to replace official
• 7:40
  national systems. Right. But people might still
• 7:43
  rely on it. You know, it really makes you think
• 7:45
  about our dependence on these automated systems
• 7:47
  and emergencies. Definitely. And, you know, speaking
• 7:50
  of systems we rely on that are changing, let's
• 7:53
  talk about something maybe a bit more every day,
• 7:55
  but still impactful. Online shopping, specifically
• 7:59
  fast fashion and shine. Prices are going up.
• 8:03
  They are. Significantly. And it's tied directly
• 8:05
  to a change in U .S. tariff policy. The de minimis
• 8:08
  exemption. That's largely gone for shipments
• 8:11
  from China. Right. The rule that let individual
• 8:13
  packages under $800 come in duty free. Exactly.
• 8:16
  Shane and Tmoo2 build a huge part of their business
• 8:19
  model on that. Sending tons of small, cheap packages
• 8:22
  directly to consumers, bypassing import duties.
• 8:26
  In 2022, something like 83 % of all U .S. e -commerce
• 8:29
  imports use that loophole. And Shane and Ti Mu
• 8:32
  made up nearly a third of that. Wow. So they
• 8:35
  had a massive advantage. But that loophole closed.
• 8:38
  It did. The Trump administration closed it for
• 8:40
  China and Hong Kong shipments back on May 2nd,
• 8:43
  2025. So now, those goods face tariffs could
• 8:47
  be 30 % on close, maybe even way higher in some
• 8:50
  cases. And we're seeing the effect already. Raiders
• 8:51
  did some price tracking. Yeah, they tracked about
• 8:53
  200 chine items from late April to late July.
• 8:56
  And the trend is pretty clear. What did they
• 8:58
  find? Well, there was a price spike right before
• 9:00
  the May 2nd deadline. Then a little dip, but
• 9:03
  prices climbed again through the summer. By late
• 9:05
  July, the average price for their cheaper stuff
• 9:07
  was up 12 .5 % compared to April. OK, 12 .5 %
• 9:11
  is noticeable. It is. And some examples were
• 9:13
  way more dramatic. They looked at 10 basic items.
• 9:16
  Their total cost went from $31 to $69 in less
• 9:18
  than three months. Whoa, over double. Yep, a
• 9:21
  123 % jump on that specific basket of goods.
• 9:24
  And it wasn't just the super cheap stuff. Higher
• 9:27
  priced items. like in the 80 to 270 dollar range,
• 9:31
  saw average prices go up about 23 percent. That's
• 9:34
  a real hit, especially for people who rely on
• 9:36
  those low prices. This must have broader implications,
• 9:39
  right? Absolutely. This isn't just about Shane's
• 9:41
  profits. Analysts point out that these price
• 9:44
  hikes hit lower income and minority communities
• 9:46
  harder, as they tend to rely more on these ultra
• 9:48
  low cost retailers. There was a National Bureau
• 9:51
  of Economic Research study that estimated ending
• 9:53
  de minimis could cost U .S. consumers somewhere
• 9:56
  between roughly 11 and 13 billion dollars overall.
• 10:00
  Billion, with a B. Billion. And poorer areas
• 10:02
  would feel more of that pain. An economist involved,
• 10:05
  Amit Khandewal from Yale, basically said it's
• 10:08
  good to raise prices, especially for lower income
• 10:09
  folks. It's a direct link between trade policy
• 10:11
  and the cost of everyday goods. So it really
• 10:14
  challenges Shane's whole approach. Cheap production,
• 10:16
  fast shipping, low friction. Totally. They're
• 10:19
  facing more delays, more costs, thinner margins.
• 10:22
  They have to pass some of that on, but they also
• 10:24
  risk losing their main appeal being cheap. Right.
• 10:27
  Tricky balance. Very. One logistics expert basically
• 10:30
  said, yeah, their margins will shrink, prices
• 10:32
  will go up. Bloomberg even reported Shane briefly
• 10:36
  cut prices in mid -May, maybe to try and keep
• 10:38
  customers after the initial tariff shock. It
• 10:41
  shows they're scrambling a bit. And this might
• 10:42
  just be the beginning, right? There's talk of
• 10:44
  even more trade actions. Yeah, the administration
• 10:46
  has set goals, potentially a full ban on de minimis
• 10:49
  by 2027. So this era of just incredibly cheap.
• 10:53
  direct from China, fast fashion. Yeah. It might
• 10:56
  genuinely be winding down for millions of shoppers.
• 10:59
  Wow. Quite a shift. OK, so wrapping up today,
• 11:02
  we've touched on AIs growing, sometimes unpredictable
• 11:04
  role in creativity, the sobering lessons from
• 11:07
  when critical tech systems falter and how global
• 11:10
  trade policy is directly changing the price of
• 11:12
  things we buy online. Yeah, it really makes you
• 11:15
  think, doesn't it? With tech evolving so fast
• 11:17
  and global rules constantly shifting, how much
• 11:20
  trust should we place in automated systems, especially
• 11:22
  for safety? And how are these bigger economic
• 11:25
  forces going to change what we can access and
• 11:27
  afford? Debra, some things to chew on. What really
• 11:30
  stood out to you today? Lots to consider. We're
• 11:33
  here to help at Snarful, so reach out on our
• 11:35
  website if you have any more questions. That's
• 11:41
  a wrap on today's top stories. Remember, tech
• 11:43
  doesn't have to be complicated, and we're here
• 11:46
  to help connect the dots. If you've got questions
• 11:49
  or want to learn more, visit us at snarful .com.
• 11:52
  Thanks for listening, and we'll catch you next
• 11:54
  time.