Latest / Connecting the Dots / Connecting the Dots - 5-19-2025

Transcript

Connecting the Dots - 5-19-2025 · Connecting the Dots

• 0:00
  Today's tech news moves fast, but understanding
• 0:03
  it shouldn't be hard. That's why Snarful Solutions
• 0:06
  Group brings you Connecting the Dots, 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. It's Monday, May 19th, 2025. We're coming
• 0:19
  to you from Sacramento, California, home base
• 0:21
  for Snarful Solutions Group. Welcome to Connecting
• 0:24
  the Dots. I'm Alex. And I'm Morgan. So before
• 0:26
  we jump in, just a quick heads up on a couple
• 0:28
  of things happening. Locally, here in Sacramento
• 0:31
  in the Central Valley, there's a red flag warning.
• 0:34
  That's active until 8 p .m. tonight. pretty critical
• 0:37
  fire weather conditions, so something to be aware
• 0:39
  of. Yeah, definitely. And looking further east,
• 0:43
  the central and southern plains right into the
• 0:44
  Ozarks, they're really getting hit with some
• 0:47
  severe weather, heavy rain, and the potential
• 0:50
  for some serious storms. We're talking strong
• 0:53
  tornadoes, large hail, damaging winds. It sounds
• 0:56
  pretty rough out there. Right. And quickly on
• 0:57
  the markets, just looking back at Friday's close
• 0:59
  May 16th, the Dow Jones finished at 42654 .74.
• 1:04
  And the S &P 500 closed up at 5958 .38. OK, so
• 1:08
  today on Connecting the Dots, we're going to
• 1:10
  do our deep dive. We've picked up three pretty
• 1:12
  interesting tech news stories. On the surface,
• 1:14
  maybe they don't seem linked. Not immediately,
• 1:16
  no. But we think if you dig a bit, you'll start
• 1:19
  to see some connections and, importantly, what
• 1:22
  they might mean for, well, for you, for everyday
• 1:24
  life. We're looking at AI in schools. a big admission
• 1:29
  from an AI research leader, and a bit of a shakeup
• 1:32
  in sports broadcasting. Sounds good. Where do
• 1:34
  we start? Let's start with Miami -Dade schools.
• 1:37
  So Miami -Dade County Public Schools, huge district,
• 1:39
  third largest in the U .S., actually. Yeah, massive.
• 1:42
  And they're making a really significant move
• 1:44
  pushing AI into their education system. That's
• 1:48
  right. It's quite something. They've already
• 1:50
  trained over 1 ,000 educators. And get this,
• 1:53
  they've given access to Google Gemini chat bots
• 1:56
  to more than 105 ,000 high school students. Wow.
• 1:59
  105 ,000. That's got to be the biggest rollout
• 2:02
  like this in the country, right? It is, yeah.
• 2:04
  Largest deployment of its kind in any US school
• 2:06
  district. And it's such a turnaround. I mean,
• 2:08
  wasn't it just like two years ago that districts,
• 2:12
  Miami included, were actually blocking these
• 2:14
  AI chat bots? Exactly. Total 180. The big worries
• 2:16
  back then were, you know, Cheating misinformation
• 2:19
  spreading all that stuff. So what changed why
• 2:22
  the The big shift now. Well, it seems like they're
• 2:26
  really looking ahead the districts talking a
• 2:27
  lot about preparing students for you know future
• 2:30
  jobs Recognizing AI is just going to be there,
• 2:33
  right? Can't ignore it. Can't ignore it. And
• 2:35
  they also want students to learn how to, like,
• 2:38
  critically assess these tools, use them responsibly.
• 2:41
  We actually heard from Roberto Gialanso. He's
• 2:44
  on the Miami Dade School Board. Oh, yeah. What
• 2:45
  did he say? He basically said, and I'm quoting
• 2:48
  here, every student should have some level of
• 2:50
  introduction to AI because it's going to impact
• 2:52
  all of our lives. Makes sense. One way or another
• 2:55
  in the tools we are using in our jobs. So that
• 2:59
  really sums up their thinking, I think. Yeah,
• 3:01
  it's a direct acknowledgement, isn't it, that
• 3:03
  AI isn't just some techie thing. It's going to
• 3:05
  be everywhere. And it's not just talk, right?
• 3:07
  There was that student example. Right. Ashley
• 3:10
  Asado, a 17 -year -old student, she mentioned
• 3:12
  using it and said that AI did a very good job
• 3:15
  of impersonating JFK. Wow. Impersonating JFK.
• 3:20
  That gives you a real picture, doesn't it? Yeah.
• 3:22
  How it's actually being used, like right now
• 3:24
  in class. Definitely makes it tangible. So thinking
• 3:27
  bigger picture. What could this kind of initiative
• 3:30
  mean down the line? Well, fundamentally, it could
• 3:33
  really change how students learn, couldn't it?
• 3:35
  You imagine a tool that helps personalize things,
• 3:38
  maybe generates practice quizzes just for you,
• 3:40
  or like that JFK example brings history to life,
• 3:43
  maybe? Yeah, making it more engaging. Exactly.
• 3:45
  And then there's the whole college and career
• 3:47
  prep side, you know, being comfortable with AI,
• 3:49
  knowing how to use it, that could be a real advantage.
• 3:52
  So it's becoming like a basic literacy, like
• 3:56
  reading and writing. but for AI. I think that's
• 3:59
  where it's heading. Yeah. This idea of AI literacy,
• 4:02
  just knowing how to handle these tools, understand
• 4:04
  their outputs, use them effectively. It's looking
• 4:07
  more and more like an essential skill for, well,
• 4:10
  for everyone navigating the world today. OK,
• 4:13
  let's shift gears a bit. Story number two. This
• 4:15
  one comes from deep inside the AI research world.
• 4:18
  We're talking about Dario Amadei. Right, the
• 4:20
  CEO of Anthropic. Yeah, Anthropic, a really big
• 4:23
  name in AI development. Yeah. And Business seems
• 4:26
  to be going well for them. Reports say they're
• 4:28
  on track for, what, $2 billion in annual revenue?
• 4:31
  That was the figure as of April, yeah, $2 billion.
• 4:35
  They've definitely made a name for themselves,
• 4:37
  especially pushing this idea of making AI that's
• 4:41
  helpful, honest, and harmless. That's their tagline.
• 4:44
  Their AI assistant, Claude, right? We hear a
• 4:46
  lot about Claude. Yeah, Claude's getting a lot
• 4:48
  of traction. But the really striking thing...
• 4:51
  For me, anyway, it was something Amadei himself
• 4:53
  said recently. A bit of a surprising admission,
• 4:55
  actually. You mean him saying that? That nobody
• 4:59
  really knows how AI works? Is that what you're
• 5:01
  referring to? That's the one. Yeah. It definitely
• 5:03
  makes you pause, doesn't it? It really does.
• 5:04
  I mean, here's the CEO of a leading AI company
• 5:07
  basically saying, look, at a really detailed
• 5:09
  level, we don't fully grasp what's going on inside
• 5:12
  these systems, even the people building them.
• 5:14
  Exactly. He used that analogy, didn't he? Saying
• 5:16
  we need something like an MRI on AI. Yeah, to
• 5:18
  see inside. Right. And he wrote this essay on
• 5:21
  his website explaining it further, basically
• 5:24
  saying that current AI, like the ones making
• 5:26
  images or writing text, they work mostly by spotting
• 5:30
  patterns in huge data sets. Not from like first
• 5:33
  principles of intelligence? Not really, no. Not
• 5:37
  like how we think humans understand things. And
• 5:39
  he had this quote, you probably saw it too. The
• 5:42
  one about unprecedented in the history of technology.
• 5:44
  That's it. This lack of understanding is essentially
• 5:47
  unprecedented, and he added that people who are
• 5:50
  worried about this ignorance, they're right to
• 5:52
  be concerned. Wow, that's that's pretty direct.
• 5:56
  Why frame it so strongly, do you think? Well,
• 5:59
  you have to remember Anthropics backstory a bit.
• 6:01
  Amadei and his sister Daniela they left open
• 6:04
  AI a few years back right over safety concerns
• 6:06
  Yeah part of it was this worry that you know
• 6:09
  the rush for progress maybe for profit was kind
• 6:12
  of pushing safety down the priority list so for
• 6:14
  Anthropic this whole lack of understanding thing.
• 6:16
  It's central to their mission They're trying
• 6:18
  to build safer AI and a big part of that is actually
• 6:21
  figuring out how these things tick And they're
• 6:23
  actively working on it, this AI interpretability
• 6:27
  idea. Yeah. Try to peek under the hood. They
• 6:29
  are. They did this interesting experiment, a
• 6:31
  red team versus blue team setup. OK, break that
• 6:34
  down. What does that mean? So imagine one team,
• 6:37
  the red team, deliberately puts a subtle flaw,
• 6:40
  like a hidden bias or a tendency, into an AI
• 6:43
  model. They kind of misalign it slightly. OK,
• 6:46
  like a hidden bug. Sort of. Then you have other
• 6:48
  teams, the blue teams, whose job is to find out
• 6:51
  what's wrong. And the key thing was, some of
• 6:53
  these blue teams used these interpretability
• 6:56
  tools techniques designed to understand the AI's
• 6:59
  internal logic. And they found the problem. They
• 7:01
  successfully identified the issue, yeah. Which
• 7:03
  suggests, you know, we are making some headway
• 7:06
  in understanding the decision -making inside
• 7:08
  these black boxes. Okay, so there's progress.
• 7:10
  But it still sounds like Amadei and others like
• 7:13
  the AI grid are waving some pretty big caution
• 7:16
  flags. The more advanced AI gets, the harder
• 7:19
  it is to understand. That's the crux of it, yeah.
• 7:21
  There's a lot of concern about these so -called
• 7:23
  emergent behaviors. Things the AI wasn't specifically
• 7:27
  programmed to do. Exactly. Stuff like deception.
• 7:30
  Or maybe even power -seeking behavior. That just
• 7:33
  - kind of emerges from the complexity and when
• 7:36
  you hear some experts saying AGI artificial general
• 7:39
  intelligence you know human level AI could potentially
• 7:41
  be here as early as 2027 2027 wow yeah it really
• 7:46
  highlights how urgent it is to figure out how
• 7:49
  to understand and frankly control these systems
• 7:52
  this lack of deep understanding makes things
• 7:54
  like regulation ethical oversight It makes it
• 7:57
  really, really challenging, especially when you
• 7:59
  think about using AI in critical areas like healthcare
• 8:02
  or finance. So stepping back, what does this
• 8:06
  mean for us, you know, just living our lives?
• 8:08
  This admission from a top AI guy that they don't
• 8:12
  fully get it. It's not just a problem for researchers,
• 8:13
  is it? No, absolutely not. I mean, think about
• 8:15
  it. If the people building these systems don't
• 8:17
  completely understand them, how much confidence
• 8:19
  can we have in their safety and reliability,
• 8:22
  especially as they get embedded deeper and deeper
• 8:24
  into everything? Like AI helping with medical
• 8:26
  diagnoses or financial decisions, self -driving
• 8:29
  cars? All of that. We need to know these things
• 8:32
  are developed responsibly, ethically, and that
• 8:35
  kind of requires knowing what's actually going
• 8:36
  on inside them. It brings a lot of questions
• 8:38
  about trust and safety to the forefront. OK,
• 8:41
  third story. We're shifting gears again quite
• 8:46
  dramatically this time moving from AI labs to
• 8:49
  The world of sports media Gary Lineker. Ah Yes
• 8:55
  match of the day exactly a huge name host of
• 8:58
  BBC's match of the day for what 25 years quarter
• 9:00
  of a century Yeah, an absolute institution for
• 9:03
  UK football fan. Well, he's announced he's stepping
• 9:05
  down as host after the 2026 World Cup end of
• 9:08
  an era really it really is and uh interesting
• 9:12
  timing perhaps in a recent interview he sort
• 9:14
  of hinted it wasn't maybe a hundred percent his
• 9:16
  call yeah i saw that he said something like he
• 9:18
  felt the bbc might have wanted him to leave yeah
• 9:21
  bring in new faces he said perhaps they wanted
• 9:23
  me to leave there was a sense of that and the
• 9:26
  bbc did announce his replacements it'll be a
• 9:28
  team approach with kelly kates mark chapman and
• 9:31
  gaby logan sharing the duties now this obviously
• 9:33
  comes after that whole thing back in 2023 tweet
• 9:35
  controversy Right. He tweeted comparing the UK
• 9:39
  government's asylum policy to 1930s Germany,
• 9:44
  which caused a huge stir. Massive. Led to him
• 9:48
  being taken off air briefly, and then the BBC
• 9:50
  had to review its whole social media policy for
• 9:53
  presenters. And he said recently he doesn't regret
• 9:55
  the content of the tweet, he still believes it
• 9:57
  was accurate. But he regretted the nonsense,
• 9:59
  as he put it. The disruption it caused the BBC.
• 10:02
  Exactly. And what I found interesting was his
• 10:04
  take on impartiality. Right. His argument is
• 10:08
  basically he's a freelancer working in sports.
• 10:11
  Why should he have the same strict impartiality
• 10:12
  rules as someone reporting the news? Yeah. He
• 10:15
  drew that distinction, and he also said he thinks
• 10:17
  the BBC worries way too much about people who
• 10:20
  dislike it instead of focusing on the millions
• 10:22
  who value it. It's a tricky line for the BBC
• 10:25
  to walk, isn't it, between allowing free expression
• 10:28
  and maintaining its required impartiality? Definitely.
• 10:31
  So even though he's leaving match of the day,
• 10:33
  he's not gone completely right. He'll still do
• 10:35
  FA Cup and World Cup coverage for them. Yeah,
• 10:37
  he's keeping those roles, and it sounds like
• 10:39
  he'll be focusing more on his podcast company.
• 10:42
  Goalhanger Podcasts. Which is actually really
• 10:44
  successful, isn't it? Shows like, the rest is
• 10:47
  politics, the rest is football. Yeah, they do
• 10:49
  very well. So he's got plenty to keep him busy.
• 10:52
  So this whole Lineker situation, what are the
• 10:55
  broader takeaways for us, you think? Well, it
• 10:57
  taps into a few big themes, doesn't it? You've
• 11:00
  got... public figures, expressing strong opinions,
• 11:03
  especially online. You've got that constant tension
• 11:05
  between free speech and impartiality, particularly
• 11:09
  for big public broadcasters like the BBC. And
• 11:12
  how those rules apply differently, maybe to different
• 11:14
  roles. Exactly. And it also just shows how the
• 11:16
  relationship between media companies and their
• 11:19
  big name talent can shift, sometimes smoothly,
• 11:22
  sometimes, well, maybe less so. Talent retention,
• 11:25
  succession planning, it's all part of that media
• 11:28
  landscape. Okay, so let's try and connect these
• 11:31
  dots quickly. We started with Miami -Dade schools,
• 11:34
  really leaning into AI for students. Quite a
• 11:37
  change from just blocking it. Right, embracing
• 11:39
  it as a tool, focusing on AI literacy. Then we
• 11:42
  went to the heart of AI development with Anthropic
• 11:45
  CEO Dario Amadei. And that really stark admission
• 11:49
  that, you know, nobody fully gets how these complex
• 11:52
  systems work yet. Big questions about safety,
• 11:55
  control, ethics. Yeah, the black box problem
• 11:58
  and the urgency around interpretability. And
• 12:00
  finally, Gary Lineker leaving match of the day,
• 12:03
  which opened up that whole can of worms about
• 12:05
  impartiality, free speech for public figures
• 12:08
  and how media institutions navigate all that.
• 12:10
  And when you put them together, you see these
• 12:12
  threads, right? The Miami story shows AI becoming
• 12:15
  mainstream, part of everyday education, raising
• 12:18
  the need for that literacy we talked about. While
• 12:20
  the Enthropic story highlights the sheer pace
• 12:22
  of AI development and the critical need to understand
• 12:25
  what we're building before it gets too advanced
• 12:26
  or integrated. Exactly. And the Lineker story,
• 12:29
  while different on the surface, reflects those
• 12:31
  ongoing debates about how we handle information,
• 12:33
  opinion, and responsibility in this digitally
• 12:36
  connected, often polarized world issues that
• 12:39
  AI will likely amplify, not simplify. They all
• 12:42
  touch on how technology is reshaping learning
• 12:44
  risk and public discourse. And that wraps up
• 12:47
  our deep dive for today. Remember, we are here
• 12:50
  to help here at Snarful, so reach out on our
• 12:52
  website if you have any more questions. That's
• 12:57
  a wrap on today's top stories. Remember, tech
• 13:00
  doesn't have to be complicated and we're here
• 13:03
  to help connect the dots. If you've got questions
• 13:06
  or want to learn more, visit us at snarful .com.
• 13:09
  Thanks for listening and we'll catch you next
• 13:11
  time.