Latest / Connecting the Dots / AI Oversight, Chip Supply Chains, and Workforce Transformation

Transcript

AI Oversight, Chip Supply Chains, and Workforce Transformation · 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. Imagine having to wait for a government
• 0:18
  agency like, say, the FDA to... Thoroughly test
• 0:24
  your smartphone's newest software update before
• 0:26
  you could actually download it. Oh, wow. Yeah,
• 0:28
  that sounds like a nightmare. Right. It sounds
• 0:30
  a little crazy, like it would just, you know,
• 0:33
  break the entire Internet. Obliquely. Well, depending
• 0:35
  on what kind of software we're talking about,
• 0:36
  that might be exactly the reality we're heading
• 0:39
  toward. Welcome to Connecting the Dots, broadcasting
• 0:42
  directly from Snarfl Solutions Group right here
• 0:44
  in Sacramento, California. I'm your host, Alex,
• 0:46
  and I'm joined by my fantastic co -host, Morgan.
• 0:49
  Hey, everyone. It's great to be here. Today is
• 0:52
  Tuesday. May 5, 2026, and we have a whole lot
• 0:56
  to unpack for you. We really do. But before we
• 0:59
  get into the heavy tech shifts happening today,
• 1:01
  let's get you caught up on what's happening outside
• 1:03
  your window and in your wallet. Yes, please.
• 1:06
  So locally here in Sacramento, you're looking
• 1:08
  at a beautiful day. It's partly sunny, sitting
• 1:12
  at a comfortable 68 degrees with just a light
• 1:15
  five mile per hour breeze. Not bad at all. Not
• 1:17
  bad, but. If you're traveling or listening from
• 1:20
  outside the valley though, it is a very busy
• 1:22
  weather day nationally. There are currently 294
• 1:26
  active U .S. weather alerts. Yeah, with 94 of
• 1:30
  those over land. So, depending on where you are
• 1:32
  right now, make sure you check weather .gov to
• 1:34
  stay safe. Absolutely. And looking at the financial
• 1:38
  markets, if you have traditional investments,
• 1:40
  it was a bit of a rough ride to start the week.
• 1:42
  Ouch, yeah. The Dow took a significant hit yesterday,
• 1:46
  closing down almost an entire percent specifically,
• 1:49
  a drop of 474 .60 points. That hurts. It does.
• 1:54
  And the S &P 500 mirrored that downward trend,
• 1:57
  dropping 27 .63 points. But, and this is fascinating,
• 2:01
  if you contrast that traditional market slide
• 2:03
  with the crypto space, it's like a completely
• 2:05
  different universe right now. Oh, Bitcoin is
• 2:07
  just on fire. It is. Bitcoin is having an absolutely
• 2:09
  incredible week. It shot up over 11 ,113 points
• 2:13
  just today, closing at a massive 80 ,937 .36.
• 2:18
  It really is a tale of two entirely different
• 2:20
  economies right now. And just a really quick
• 2:22
  update from the Diamond. If you caught the game
• 2:24
  last night, the SF Giants edged out the San Diego
• 2:27
  Padres with a tight 3 -2 victory. Yeah, big shout
• 2:31
  out to winning pitcher Trevor McDonald for holding
• 2:33
  the line in those final innings. And both the
• 2:36
  Oakland Athletics and our hometown Sacramento
• 2:38
  Kings had the night off. So, you know, a rare
• 2:41
  chance to reclaim your evening. Love a quiet
• 2:43
  sports night sometimes. All right. So there is
• 2:46
  a massive wave of technological shifts happening
• 2:48
  right under our noses today. Let's look closer.
• 2:51
  Today's episode is brought to you by Strong Coffee.
• 2:53
  The premium instant coffee that tastes like a
• 2:55
  traditional brew but is packed with 15 grams
• 2:58
  of protein, MCTs, and adaptogens like ashwagandha.
• 3:02
  Get clean, sustained energy without the jitters.
• 3:04
  Use promo code SNARFL for 20 % off at strongcoffeecompany
• 3:08
  .com. Okay, let's jump right into our Morningstar
• 3:11
  tech report. Today, our three main stories might
• 3:14
  initially seem like they're, I don't know, pulled
• 3:17
  from completely different corners of the business
• 3:19
  world. Right, they seem disconnected at first
• 3:21
  glance. Exactly. But they're actually all anchored
• 3:24
  by a single massive force. We're looking at the
• 3:28
  immediate real world ripple effects of the AI
• 3:31
  explosion. It's touching everything. It really
• 3:34
  is. We're going to trace this from how the government
• 3:36
  is actively trying to regulate AI software to
• 3:39
  how the physical hardware supply chain is just
• 3:42
  it's buckling under the weight. of those models.
• 3:45
  Oh, big time. And finally, to how this technology
• 3:48
  is fundamentally rewriting what it means to even
• 3:51
  hold a job. It really is the ultimate domino
• 3:53
  effect. And the first domino falling right now
• 3:57
  is government regulation. Right. So what's happening
• 3:59
  in Washington? Well, according to reports from
• 4:01
  the New York Times, Wall Street Journal, and
• 4:03
  Axios, the Trump administration is currently
• 4:06
  in early exploratory discussions about forming
• 4:09
  a new AI working group. Okay. And the core focus
• 4:12
  of this group would be establishing strict oversight
• 4:15
  procedures for advanced AI models. Specifically,
• 4:18
  they're looking at vetting these high -risk models
• 4:20
  before they are ever released to the public.
• 4:23
  And just to be clear for everyone listening,
• 4:25
  when we say high risk, we aren't talking about
• 4:27
  like a chat bot giving you a hallucinated recipe
• 4:30
  for chocolate chip cookies. No. We're talking
• 4:33
  about actual national security level threats.
• 4:36
  Exactly. The conversations at the federal level
• 4:38
  are heavily focused on AI systems that are capable
• 4:41
  of facilitating massive cyber attacks. To understand
• 4:45
  why this is such a big deal, you kind of have
• 4:47
  to look at how these models actually work. Right.
• 4:49
  They aren't just writing simple scripts anymore.
• 4:51
  Not at all. An advanced AI can digest the entire
• 4:54
  code base of, say, a hospital's legacy computer
• 4:57
  network, understand the underlying logic, and
• 5:00
  then actively help malicious users identify zero
• 5:03
  -day vulnerabilities. Wow. So it spots the holes
• 5:06
  in the fence that human security teams haven't
• 5:08
  even noticed yet. Precisely. So officials are
• 5:11
  weighing several options to stop this, which
• 5:13
  could include formal mandatory pre -release review
• 5:17
  processes or even government -led testing environments
• 5:20
  for these specific higher risk systems. To give
• 5:23
  you some context on the mechanics of this policy
• 5:26
  shift, this actually really mirrors the multi
• 5:29
  -layered oversight approach we're currently seeing
• 5:32
  over in the UK. Right, where they essentially
• 5:34
  require confirmation that models meet certain
• 5:36
  safety standards before commercial deployment.
• 5:39
  Exactly. And it's also worth noting the timeline
• 5:41
  here. On January 20, 2025, which was his first
• 5:44
  day back in office, Donald Trump officially revoked
• 5:47
  the Biden administration's executive order 14
• 5:50
  -110 on AI. Which was a pretty major shift. Huge.
• 5:54
  But what we're seeing now is simply the mechanical
• 5:57
  process of a new administration trying to build
• 5:59
  its own regulatory framework from the ground
• 6:01
  up. Right. And regardless of the political mechanics
• 6:04
  in Washington, the fundamental practical challenge
• 6:06
  remains exactly the same. Like, how do you regulate
• 6:09
  a piece of software that evolves faster than
• 6:12
  the legislative process can even draft a bill.
• 6:14
  Seriously, they're so slow. Right. If a formal
• 6:16
  oversight group is created and they actually
• 6:18
  implement mandatory pre -release testing, it
• 6:21
  represents a massive tectonic shift in how the
• 6:24
  tech industry operates. Yeah, for working tech
• 6:26
  professionals, this could mean significant delays
• 6:29
  in getting access to the cutting edge tools you
• 6:32
  rely on to do your job. Yep. But on the flip
• 6:34
  side, for families and everyday Internet users,
• 6:38
  it could theoretically create a much safer digital
• 6:41
  environment by keeping exploit generating AI
• 6:45
  completely out of the wild. I'm just trying to
• 6:47
  wrap my head around what this actually looks
• 6:49
  like in practice for a developer. Like if we
• 6:52
  start demanding that. software be rigorously
• 6:55
  tested and approved by a government body before
• 6:57
  it can be sold or downloaded? It changes everything.
• 7:00
  It feels like we're entering a totally new paradigm.
• 7:03
  Are we essentially entering an era where software
• 7:05
  requires clinical trials before the public can
• 7:08
  use it? Like is this going to be the FDA but
• 7:10
  for lines of code? It's a really interesting
• 7:13
  analogy but I'd argue it's actually much more
• 7:15
  complicated than the FDA. Oh really? How so?
• 7:18
  Well think about a pharmaceutical clinical trial.
• 7:20
  Before a company can sell a new heart medication,
• 7:23
  they have to prove it's safe and effective. It
• 7:25
  takes years and costs millions. But here's the
• 7:28
  key difference. A pill's chemical formula is
• 7:31
  static. It doesn't change once it leaves the
• 7:33
  factory. Software is incredibly dynamic. The
• 7:37
  entire tech industry has historically lived by
• 7:41
  the mantra of move fast and break things. Right.
• 7:43
  You release the software, let the users find
• 7:45
  the bugs, and you just patch it with an update
• 7:47
  over the weekend. Exactly. But you can't exactly
• 7:50
  push a quick weekend patch to fix a cyber attack
• 7:53
  that just took down a city's power grid. Yeah,
• 7:56
  that's a fair point. When the thing being broken
• 7:59
  is critical infrastructure, the beta testing
• 8:01
  mindset completely falls apart. Precisely. But
• 8:05
  returning to your FDA analogy, an AI model isn't
• 8:07
  just dynamic, it learns. You could have a government
• 8:10
  working group test a model on a Tuesday and certify
• 8:13
  that it's safe. But by Thursday, after interacting
• 8:16
  with a million users, That exact same model might
• 8:19
  have figured out a new, creative way to generate
• 8:21
  malicious code. Wow, so the test is instantly
• 8:23
  obsolete. Exactly. If this pre -release vetting
• 8:27
  becomes a reality, tech companies won't just
• 8:29
  have to build a clinical trial phase into their
• 8:32
  development lifecycle. They'll have to figure
• 8:34
  out how to prove that a learning machine won't
• 8:36
  suddenly unlearned at safety guardrails. That
• 8:39
  sounds incredibly difficult. It is. It would
• 8:41
  completely bottleneck the speed of innovation,
• 8:43
  which is why Silicon Valley is watching these
• 8:45
  White House discussions with so much anxiety.
• 8:48
  Well, so while the government is trying to figure
• 8:50
  out how to put a speed limit on the software
• 8:53
  side of AI, the tech giants might hit a wall
• 8:57
  anyway. Oh, they definitely are. Because ironically,
• 8:59
  the physical hardware required to actually run
• 9:01
  these massive models is completely buckling.
• 9:04
  Yes, the physical strain is immense and it's
• 9:07
  hitting the biggest company on earth right now.
• 9:09
  Apple. Apple. According to Bloomberg, Apple has
• 9:12
  been holding exploratory talks with both Intel
• 9:15
  and Samsung to potentially manufacture the core
• 9:18
  processors for Macs and iPhones right here in
• 9:21
  the United States. Really? Yeah. Apple executives
• 9:24
  have even visited a massive Samsung fabrication
• 9:26
  plant that's currently under construction in
• 9:29
  Taylor, Texas. Now, to understand why Apple flirting
• 9:32
  with other manufacturers is such a massive deal,
• 9:35
  we have to look at how they build their devices
• 9:37
  today. Right, because right now, they're heavily
• 9:40
  dependent. Almost entirely. Right now, Apple
• 9:43
  relies almost entirely on one single company
• 9:45
  for its most advanced chips. TSMC, the Taiwan
• 9:48
  Semiconductor Manufacturing Company. Apple desperately
• 9:52
  wants to diversify, and they're being pushed
• 9:54
  by two massive forces. Right, the first being
• 9:56
  the obvious geopolitical risk surrounding Taiwan.
• 9:59
  Exactly. But the second, and the one causing
• 10:02
  immediate operational pain today, is the AI -driven
• 10:04
  chip shortage. It is a supply chain traffic jam
• 10:07
  of historic proportions. Like, the entire tech
• 10:11
  industry is frantically building out massive
• 10:13
  AI data centers, and every single one of those
• 10:16
  servers needs advanced silicon. Didn't Nvidia's
• 10:19
  CEO just say something about this? Yeah, Jensen
• 10:21
  Huang recently mentioned that Nvidia has actually
• 10:24
  dethroned Apple as TSMC's number one customer.
• 10:26
  That is wild. It is. Because so much of the global
• 10:30
  manufacturing capacity is being eaten up by AI
• 10:33
  server chips, Apple is really feeling the squeeze
• 10:36
  on its consumer products. Just last week, Apple
• 10:39
  CEO Tim Cook admitted on an earnings call that
• 10:42
  they are facing major supply constraints specifically
• 10:45
  for the Mac Mini and the Mac Studio. And it is
• 10:48
  absolutely not a coincidence that those specific
• 10:50
  computers are the ones facing constraints. If
• 10:53
  you're a developer or a power user, the Mac Mini
• 10:56
  and the Mac Studio are the exact machines you
• 10:59
  buy to run local open source AI models right
• 11:02
  on your desk without having to pay for expensive
• 11:05
  cloud computing time. It's a massive localized
• 11:08
  AI boom. Yeah, and Cook basically admitted it
• 11:10
  could take several months just to balance out
• 11:12
  the supply and demand. So Apple is aggressively
• 11:14
  looking for backup foundries. And historically,
• 11:17
  you know, this is an entirely new territory for
• 11:19
  them. Right. They've partnered up before. Exactly.
• 11:22
  Yeah. From the very first iPhone all the way
• 11:25
  up to the iPhone 5S. Apple actually used Samsung
• 11:29
  chips. And of course, if you owned a Mac before
• 11:32
  2023, you remember the Intel inside era before
• 11:36
  Apple transitioned to their own custom M series
• 11:39
  Silicon architecture. I remember my old Intel
• 11:41
  MacBook getting so hot it could cook an egg.
• 11:43
  Same. But the landscape has completely changed
• 11:46
  since then Apple designs the chicks themselves
• 11:49
  now But they still need someone to physically
• 11:51
  build them at a microscopic scale and TSMC is
• 11:54
  currently the undisputed king of that specific
• 11:57
  manufacturing process Okay, let's unpack this
• 11:59
  from your perspective the person listening who
• 12:01
  might just be walking into an Apple store to
• 12:02
  buy a new computer for work or school. If Apple
• 12:05
  successfully dual sources these chips, meaning
• 12:08
  some iPhones get a processor built by TSMC in
• 12:11
  Taiwan, and some get a processor built by Samsung
• 12:14
  or Intel in Texas, I have a major concern about
• 12:17
  component parity. I know exactly where you're
• 12:19
  going with this. We've seen this exact movie
• 12:21
  before in the tech world. Do you remember when
• 12:24
  Apple used both LG and Samsung to manufacture
• 12:27
  iPhone screens, and users immediately started
• 12:30
  noticing a distinct difference in color quality?
• 12:32
  Oh yeah, people were furious. Or over in the
• 12:35
  Android ecosystem, where phones with supposedly
• 12:38
  identical Snapdragon processors had wildly different
• 12:41
  battery lives entirely depending on if the chip
• 12:43
  was physically manufactured by TSMC or Samsung?
• 12:46
  Ah, yes. the infamous silicon lottery. It's a
• 12:49
  very real, very costly concern for hardware manufacturers.
• 12:54
  Right. So if this dual sourcing happens, are
• 12:56
  you going to have to start checking the ingredients
• 12:58
  label on the back of your new MacBook box to
• 13:00
  see if you got the good TSMC chip or the slightly
• 13:04
  less efficient backup Intel chip? To understand
• 13:07
  why that happens, you have to look at the mechanics
• 13:09
  of what TSMC actually does. They aren't just
• 13:12
  stamping out plastic toys. They are building
• 13:14
  advanced nodes. Which means what exactly? Basically,
• 13:17
  that means they are perfectly placing billions
• 13:20
  of microscopic transistors onto a tiny square
• 13:23
  of silicon. TSMC has spent decades perfecting
• 13:26
  what the industry calls yield and efficiency.
• 13:30
  Yield is just the percentage of chips on a raw
• 13:33
  silicon wafer that actually function perfectly.
• 13:35
  OK, so it's like baking a massive batch of cookies
• 13:37
  and seeing how many come out perfectly round
• 13:39
  versus like slightly burnt on the edges. Exactly.
• 13:43
  And TSMC burns very few cookies. If Intel or
• 13:46
  Samsung are brought in as backup manufacturers,
• 13:49
  they might be using a slightly different microscopic
• 13:51
  process technology to build the chip. Even with
• 13:54
  Apple's designs? Yep. Even if Apple hands them
• 13:57
  the exact same architectural blueprints for an
• 13:59
  M -series processor, the physical properties
• 14:01
  of how that resulting chip handles heat, or how
• 14:04
  efficiently it sips battery power, could vary.
• 14:07
  So your concern about an ingredients label is
• 14:10
  completely valid. Imagine you and your co -worker
• 14:12
  buy the exact same $2 ,000 laptop on the exact
• 14:15
  same day. But because yours has an Intel chip
• 14:18
  fabricated in Texas, and theirs has a TSMC chip
• 14:22
  from Taiwan, your laptop runs five degrees hotter
• 14:25
  and your battery dies two hours earlier every
• 14:29
  single afternoon. That is the ultimate Silicon
• 14:32
  Lottery nightmare. It really is. It's a massive
• 14:35
  quality control headache for a brand that has
• 14:38
  built its entire identity on perfection. It really
• 14:41
  just shows how desperate the supply chain situation
• 14:43
  has become. But here is where the ripple effect
• 14:46
  hits closest to home. Oh, yeah, this next part
• 14:48
  is big. The AI boom isn't just forcing the world's
• 14:51
  biggest company to completely rethink its physical
• 14:54
  hardware supply chain. It's forcing companies
• 14:56
  everywhere to radically rethink their human workforce.
• 15:00
  The hardware strain is ultimately leading to
• 15:02
  human redundancy. Yeah, this one is going to
• 15:04
  resonate with anyone working in the corporate
• 15:05
  world right now. Coinbase, one of the largest
• 15:08
  cryptocurrency exchanges globally, announced
• 15:10
  yesterday that they're laying off roughly 700
• 15:13
  employees. Wow. It's about 14 % of their entire
• 15:18
  global workforce. In an internal email that was
• 15:21
  shared publicly, Coinbase CEO Brian Armstrong
• 15:24
  cited two main converging forces for these cuts.
• 15:30
  That was the first one. The current down market
• 15:32
  and the broader crypto cycle. But second, and
• 15:35
  far more notably for our deep dive today, he
• 15:38
  pointed to massive, undeniable productivity gains
• 15:41
  driven entirely by AI. The details of how they
• 15:44
  are restructuring the company post layoffs are
• 15:47
  fascinating. They're changing everything. Right.
• 15:49
  They aren't just cutting headcount to save a
• 15:51
  few million dollars. They are entirely redesigning
• 15:54
  what the corporate ladder looks like. Coinbase
• 15:56
  is aggressively flattening its organizational
• 15:58
  chart. Going forward, there will be a maximum
• 16:00
  of five layers of management between the lowest
• 16:02
  level entry employee and the CEO. That's incredibly
• 16:06
  flat for a company that size. And more importantly,
• 16:08
  they are completely eliminating cure management
• 16:11
  roles. From now on, every single manager has
• 16:14
  to be what they're calling a player coach. It
• 16:16
  really signals the death of the traditional middle
• 16:19
  manager whose only daily function is to oversee
• 16:22
  other people's work, you know, track deadlines
• 16:25
  and attend meetings. Right. If you want to lead
• 16:27
  a team at Coinbase now, you also have to be an
• 16:30
  active individual contributor. You have to write
• 16:33
  code, design products, or ship tangible features
• 16:36
  while simultaneously managing up to 15 direct
• 16:39
  reports. That sounds overwhelming. It does. To
• 16:42
  make this possible, they're focusing heavily
• 16:44
  on building what they call AI native pods. These
• 16:48
  are incredibly small, highly focused teams that
• 16:51
  heavily leverage AI tools to do the work that
• 16:54
  used to require a much larger group of humans.
• 16:56
  And didn't they mention one -person teams? Yes.
• 16:59
  They're even experimenting with one -person teams,
• 17:02
  where a single employee is responsible for engineering,
• 17:05
  design, and product management all at once. Let's
• 17:07
  stop and think about how a one -person team actually
• 17:10
  functions. Like, if I'm that single employee,
• 17:12
  I might have a great idea for a product, but...
• 17:14
  Maybe I don't know how to code the backend database.
• 17:17
  Like you'd be stuck. In the past, my project
• 17:19
  would just stall until I could hire a database
• 17:21
  engineer. Now, I just prompt an AI coding agent
• 17:24
  to write the database structure for me. The AI
• 17:27
  fills in the gaps in my skill set. Exactly. Armstrong's
• 17:30
  email really spelled this out. He wrote, over
• 17:32
  the past year, I've watched engineers use AI
• 17:35
  to ship in days what used to take a team weeks.
• 17:38
  Non -technical teams are now shipping production
• 17:41
  code, and many of our workflows are being automated.
• 17:44
  And to put this Coinbase restructuring into a
• 17:47
  broader macroeconomic context, this isn't just
• 17:50
  one crypto company using AI as an excuse to make
• 17:53
  budget cuts. No, it's an industry -wide trend.
• 17:56
  Goldman Sachs economists recently released an
• 17:59
  estimate that AI substitution is currently erasing
• 18:01
  roughly 25 ,000 US jobs every single month. That
• 18:04
  number is staggering. It is. And while the AI
• 18:07
  boom is certainly creating new jobs, you know,
• 18:10
  data specialists, prompt engineers, it's only
• 18:12
  adding back about 9 ,000 jobs a month. Mathematically,
• 18:15
  that is a net loss of 16 ,000 positions monthly
• 18:18
  across the economy. We've seen similar AI -driven
• 18:20
  restructuring and layoffs recently at Meta, Microsoft,
• 18:24
  and Jack Dorsey's company, Block. I have to admit,
• 18:27
  hearing those numbers is a little terrifying.
• 18:29
  It feels like we're watching the Industrial Revolution
• 18:31
  happen in real time, but entirely for white -collar
• 18:35
  office work. That's a really good way to put
• 18:36
  it. Right. If one single person can now act as
• 18:40
  an entire department, doing the coding, the graphic
• 18:43
  design, and the project management before lunch
• 18:45
  just by prompting AI tools, Are you and I gonna
• 18:49
  have to act more like solopreneurs to survive
• 18:53
  in the modern corporate workforce? It sounds
• 18:55
  exhausting. Yeah, doing everything yourself.
• 18:57
  Exactly! You'd have to be your own HR department,
• 19:00
  your own designer, your own quality assurance
• 19:02
  tester. Well, I wouldn't necessarily say you
• 19:04
  have to become a literal solopreneur, where you
• 19:07
  work entirely alone, but you absolutely have
• 19:10
  to become the manager of your own output. Okay,
• 19:12
  what do you mean by that? The traditional corporate
• 19:14
  mindset for the last 50 years. was based on deep
• 19:18
  narrow specialization. It was an assembly line.
• 19:22
  You were the database engineer or you were the
• 19:24
  UX designer and you did your specific task and
• 19:27
  then you handed your work off to the next specialist
• 19:29
  down the hall. Yep, the classic handoff. But
• 19:31
  AI destroys the necessity of that assembly line
• 19:34
  because it gives everyone instant access to baseline
• 19:37
  competency in every single field. Right. If you're
• 19:39
  a graphic designer, an AI agent can write the
• 19:42
  backend code to make your design functional.
• 19:44
  Exactly. So to survive this shift, you can't
• 19:48
  just be a cog in the machine waiting for instructions
• 19:50
  anymore. You have to understand the entire lifecycle
• 19:53
  of a product from conception to deployment. You
• 19:57
  essentially have to see the big picture. Yeah.
• 19:59
  You bring your core expertise to the table, but
• 20:01
  you use AI to flawlessly execute the parts you
• 20:05
  aren't an expert in. It makes sense. But it's
• 20:07
  still a tough pill to swallow for the people
• 20:09
  losing their jobs right now. Definitely. For
• 20:11
  what it's worth, Coinbase is offering a very
• 20:13
  significant severance package to the people they
• 20:15
  let go. Oh, they are? Yeah. U .S. employees are
• 20:18
  receiving a minimum of 16 weeks of base pay,
• 20:21
  plus two additional weeks for every year of service,
• 20:24
  their next equity vest, and six months of health
• 20:27
  care coverage. OK, that gives them a fairly soft
• 20:29
  landing. It does. But the harsh reality is...
• 20:32
  They are stepping out into a job market that
• 20:35
  is fundamentally changing the definition of what
• 20:37
  an employee actually is. Before we wrap up, a
• 20:41
  quick thanks to Old Glory, an iconic music and
• 20:44
  sports fan merch store, perfect for the techie
• 20:46
  in your life. Use promo code SNARFL for 15 %
• 20:50
  off at oldglory .com. So if we step back and
• 20:53
  connect the dots on everything we've unpacked
• 20:55
  today, the narrative thread is pretty undeniable.
• 20:58
  Yeah, AI is at the center of it all. Artificial
• 21:00
  intelligence is no longer just a cool tool or
• 21:02
  a futuristic concept you read about. It's the
• 21:05
  physical, structural, and economic force reshaping
• 21:08
  your reality right now. Absolutely. It's sparking
• 21:10
  the highest levels of government to completely
• 21:12
  reconsider how we ensure public safety against
• 21:15
  dynamic learning software threats. It's causing
• 21:18
  a historic bottleneck in global chip manufacturing,
• 21:21
  forcing an absolute giant like Apple to rethink
• 21:24
  decades of microscopic supply chain strategy.
• 21:26
  And of course the jobs. Right. It's completely
• 21:29
  altering the definition of what it means to be
• 21:30
  an employee, collapsing entire corporate departments
• 21:33
  into single, highly empowered individuals. It
• 21:36
  really leaves you wondering about the next generation.
• 21:39
  Like, as AI models become increasingly capable
• 21:42
  of doing the heavy lifting of an entire team,
• 21:45
  will the future of education and job training
• 21:47
  need to completely shift? It almost has to, right?
• 21:50
  Yeah. Instead of spending four years in college
• 21:52
  learning a specific technical skill, like...
• 21:55
  how to write a specific coding language, will
• 21:58
  we need to be trained purely on how to be effective
• 22:00
  managers of artificial intelligence? That's a
• 22:03
  wild thought. It's a completely different way
• 22:05
  of looking at human potential, and it's a reality
• 22:07
  we're going to have to face much sooner than
• 22:09
  we originally thought. We're here to help at
• 22:11
  SNARFL, so reach out on our website if you have
• 22:13
  any more questions. That's a wrap on today's
• 22:19
  top stories. Remember, tech doesn't have to be
• 22:22
  complicated, and we're here to help connect the
• 22:24
  dots. If you've got questions or want to learn
• 22:27
  more, visit us at snarful .com. Thanks for listening,
• 22:31
  and we'll catch you next time.