Latest / Connecting the Dots / Connecting the Dots - 5-22-2025

Transcript

Connecting the Dots - 5-22-2025 · Connecting the Dots

• 0:00
  Today's tech news moves fast, but understanding
• 0:03
  it shouldn't be hard. That's why Snarfle 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. Let's get started. It is Thursday, May
• 0:17
  22nd, 2025. And welcome everyone to Connecting
• 0:22
  the Dots. Good to be here. I'm Alex. And I'm
• 0:24
  Morgan. And we are colleagues here at SNARFL
• 0:26
  Solutions Group, recording from, well, sunny
• 0:29
  Sacramento today. Yeah, it's actually really
• 0:30
  nice out, what is it, 85 degrees? About that,
• 0:33
  yeah. A perfect day. Quick look at the markets
• 0:36
  before we jump in. Yesterday, Wednesday the 21st,
• 0:40
  the Dow closed at 418 .60 .44 and the S &P 500
• 0:45
  finished up at 588 .44 .61. Got it. So today,
• 0:50
  we're going to tackle three big tech stories
• 0:53
  that really caught our eye. Yeah. And we'll try
• 0:56
  to break down not just what they are, but what
• 0:58
  they might actually mean for you. Exactly. How
• 1:00
  does this connect back? It's pretty wild how
• 1:02
  fast everything's moving, isn't it, these stories?
• 1:04
  Faster AI, Bitcoin going crazy, and even books
• 1:07
  about AI leaders. Right. They seem separate,
• 1:10
  but they really show how dynamic tech is and
• 1:15
  how much it's weaving into everything. So first
• 1:17
  up, Google. DeepMind, Gemini Diffusion, that
• 1:21
  was part of their big IO event, right? Yeah,
• 1:23
  exactly. They announced Gemini Diffusion. It's
• 1:25
  still experimental, available via waitlist, I
• 1:27
  think. Experimental, okay. But what's the headline?
• 1:30
  What's got people talking? Well, it's the speed,
• 1:34
  the text generation speed. They're claiming it
• 1:36
  can generate text at 1 ,000 to 2 ,000 tokens
• 1:40
  per second. Whoa, okay. One thousand to two thousand.
• 1:44
  Yeah tokens per second put that in perspective
• 1:46
  for us. Like what's normal? Well normal varies
• 1:49
  but think about gruck. They're known for being
• 1:51
  super fast, right? They're around like 275 tokens
• 1:55
  per second and your typical providers are Quite
• 1:59
  a bit slower than even that this is a massive
• 2:02
  giant huge leap Yeah, a token is roughly, you
• 2:05
  know a word or part of a word. So this is just
• 2:07
  different league how how does it work so differently
• 2:10
  I thought AI text was usually like one word and
• 2:13
  the next word sequentially that's the standard
• 2:15
  way yeah it's called autoregressive it predicts
• 2:18
  the next bit based on the previous bits like
• 2:20
  your phone keyboard guessing the next word diffusion
• 2:23
  models which we've seen work really well for
• 2:26
  image generation they do it differently they
• 2:28
  sort of start with randomness like digital static
• 2:30
  almost and then they refine that noise step -by
• 2:34
  -step based on the prompt until coherent text
• 2:37
  emerges. It's like focusing a blurry picture.
• 2:40
  So less like building, more like sculpting. That's
• 2:43
  a good way to put it. And yeah, Gemini Diffusion
• 2:45
  is really the first big attempt to apply this
• 2:48
  image technique to text generation on this scale.
• 2:51
  And people who've seen it are impressed. I heard
• 2:52
  jaw dropping. Yeah, apparently the responsiveness
• 2:55
  and demos is pretty remarkable. Near real time,
• 2:59
  it fit right in with Google's whole IO theme
• 3:01
  this year. Lots of AI upgrades focus on speed
• 3:04
  integration. OK, so let's connect the dots for
• 3:07
  us here at Starfall and for everyone listening.
• 3:10
  Beyond just being cool tech, what does this mean
• 3:12
  practically? What changes? Well, the potential
• 3:14
  is pretty significant. Faster text generation
• 3:17
  means potentially much quicker AI assistance.
• 3:20
  Like ask a question, get an answer instantly.
• 3:22
  Almost, yeah. Or think about content creation
• 3:24
  tools, drafting emails, reports. Imagine that
• 3:28
  happening in true real time. OK, that could speed
• 3:31
  things up. For us internally, maybe faster data
• 3:35
  analysis, quicker comms, as this stuff gets baked
• 3:37
  into the software we all use, you could see real
• 3:40
  efficiency gains. Definitely something to watch.
• 3:42
  Yeah. Okay. Let's shift gears completely. Bitcoin.
• 3:47
  Ah, yes. It never stays quiet for long, does
• 3:50
  it? Not really, it's hitting new records again.
• 3:52
  Yeah, big time. It's smashed past its old highs,
• 3:55
  went over $109 ,000, then $109 ,486 yesterday.
• 4:00
  And today, Thursday, it's actually crossed $110
• 4:03
  ,000. Wow, 110k, what kind of jump is that from
• 4:07
  recent lows? It's surged over 46 % just since
• 4:11
  the low point back in April. 46%. Okay, what's
• 4:14
  fueling this fire? Is it just hype again? Well,
• 4:16
  analysts think it might be a bit more sustainable
• 4:18
  this time. There are a few key drivers. Public
• 4:20
  what? First, those spot Bitcoin ETFs, they're
• 4:22
  seeing huge inflows again in May, about $3 .6
• 4:25
  billion net. So more mainstream money coming
• 4:27
  in through those funds. Exactly. It makes it
• 4:29
  easier for people to get exposure. Second, you've
• 4:31
  got these big Bitcoin -focused companies like
• 4:33
  Michael Saylor's Strategy and this new one, 21
• 4:36
  Capital. They're actively buying more Bitcoin.
• 4:40
  That signals confidence. So... Serious players
• 4:43
  doubling down. Right. And then there's the regulatory
• 4:45
  side. Things are looking, well, maybe slightly
• 4:49
  more positive in the U .S. How so? There's a
• 4:51
  stablecoin bill making progress in the Senate.
• 4:54
  Little steps like that seem to add legitimacy
• 4:56
  or at least reduce uncertainty for digital assets.
• 4:59
  Okay, that makes sense. Anything else? broader
• 5:02
  market stuff. Yeah, could be. Maybe easing trade
• 5:05
  tensions help risk appetite. And even Moody's
• 5:08
  downgrading US debt might push some investors
• 5:11
  to look at alternatives. Huh. Interesting link.
• 5:14
  Plus the Nasdaq is doing well, which often means
• 5:17
  investors are feeling good about tech and growth,
• 5:19
  maybe spilling over into crypto. And didn't I
• 5:21
  see something about Jamie Dimon? He used to be
• 5:23
  super anti -Bitcoin. That's a big signal, yeah.
• 5:25
  Even Jimmy Diamond, who was famously skeptical,
• 5:28
  now says JP Morgan will let clients buy Bitcoin.
• 5:31
  Wow. And Coinbase just got added to the S &P
• 5:33
  500. So traditional finance is definitely getting
• 5:36
  more involved. So the thinking is, yeah, this
• 5:38
  isn't just froth. It's got legs this time. That
• 5:40
  seems to be the analyst consensus. Yeah. Better
• 5:43
  financial conditions, ETF flows, institutional
• 5:46
  buying. Maybe less wild speculation than before.
• 5:50
  OK, so for people listening who maybe don't own
• 5:53
  Bitcoin, why should they care about this rally?
• 5:55
  How does it connect? Well, even if you're not
• 5:57
  invested, Bitcoin's price can be barometer for
• 6:03
  overall risk appetite in the market. Like a confidence
• 6:06
  indicator? Sort of, yeah. When investors are
• 6:09
  piling into something seen as risky, like Bitcoin,
• 6:12
  it might suggest broader optimism, which could
• 6:14
  affect your stocks, your 401K, the whole economy,
• 6:17
  potentially. OK. And just the fact that it's
• 6:19
  becoming more integrated with traditional finance
• 6:21
  means it's, well, it's becoming part of the financial
• 6:24
  furniture we all need to be aware of. Gotcha.
• 6:26
  Good perspective. All right. Third topic, shifting
• 6:28
  back to AI, but more about the people and the
• 6:30
  narrative. Right. Specifically, Sam Altman and
• 6:33
  OpenAI. There are two new books out about them.
• 6:35
  Two, okay. Do they agree? Not entirely, no. They
• 6:39
  offer pretty different takes. One is Empire of
• 6:42
  AI. by Karen Howe. And that one's, let's say,
• 6:45
  more critical. It compares the AI industry's
• 6:49
  hunger for resources data, computing power, to,
• 6:52
  like, historical colonialism. Ooh, strong comparison.
• 6:56
  Yeah. It suggests this race for artificial superintelligence
• 7:00
  is maybe becoming this all -consuming thing with
• 7:03
  hidden costs. A cautionary tale. Sort of. That's
• 7:07
  the other book. It's called The Optimist by Keach
• 7:09
  Hagee. And while the title sounds positive, the
• 7:11
  book apparently leaves you wondering if Altman
• 7:13
  himself, the sort of unassuming figure, might
• 7:16
  end up wielding enormous power through AI. Interesting.
• 7:19
  So one worries about the process, the other maybe
• 7:22
  about the person at the center. Kind of, yeah.
• 7:24
  Different angles on the whole phenomenon. It's
• 7:26
  fascinating to see these deep dives happening
• 7:28
  while everything is still unfolding so fast.
• 7:30
  Absolutely. And Altman himself had some observations
• 7:33
  recently about how people are actually using
• 7:35
  these tools. like ChatGPT. Oh yeah, what did
• 7:37
  he say? He pointed out a generational difference.
• 7:39
  Younger users, Gen Z, Millennials, especially
• 7:43
  college students, they seem to be using it more
• 7:45
  like a life advisor. The life advisor, seriously?
• 7:47
  Yeah. Or even an operating system for parts of
• 7:50
  their life, integrating it really deeply into
• 7:52
  decisions, workflows. Wow, compared to older
• 7:56
  users. Who might use it more like a better search
• 7:59
  engine. for specific tasks, information retrieval.
• 8:03
  That's a really stark difference in adoption.
• 8:05
  Yeah. Using AI as your OS. It definitely raises
• 8:08
  questions, right? Yeah. About trust, reliance,
• 8:11
  the safety of taking life a device from an LLM.
• 8:14
  Experts are debating the ethics of that. For
• 8:16
  sure. Okay, and there's one more quirky thing.
• 8:18
  The chatbots themselves have opinions on Albin.
• 8:22
  Yeah, that was funny. Someone prompted a bunch
• 8:24
  of the big AI chatbots. Chat, GPT, Claude, Copilot,
• 8:29
  Gemini, Meta, AI, Perplexity, asking who should
• 8:32
  lead AI development, Sam Altman or Elon Musk?
• 8:35
  No, what did they say? Most of them leaned towards
• 8:37
  Altman. They mentioned things like his focus
• 8:39
  on safety, collaboration, his track record. Interesting.
• 8:42
  What about Musk's own chat bot, Grok? Predictably,
• 8:45
  Grok picked Musk. It cited his focus on, quote,
• 8:48
  long -term survival and first principles thinking.
• 8:51
  Of course it did. It's like asking a Ford who
• 8:53
  makes the best truck. Pretty much, but it does
• 8:55
  kind of highlight the different philosophies
• 8:57
  out there. So wrapping this section up, what's
• 9:00
  the main takeaway for you from these books, the
• 9:02
  usage patterns? How does this connect? I think
• 9:04
  it just underscores how deeply AI is starting
• 9:08
  to permeate things from big picture questions
• 9:11
  about leadership and control down to how you
• 9:14
  personally might use it for advice or work. Right.
• 9:17
  Understanding these different views and seeing
• 9:19
  how usage is evolving differently across generations
• 9:22
  helps you think more critically about AI's impact
• 9:25
  now and where it might be heading. Makes sense.
• 9:29
  It's not just code, it's how we interact with
• 9:31
  it. Exactly. Okay, so let's quickly recap the
• 9:33
  big points from today's deep dive. Sounds good.
• 9:36
  We started with Google's Gemini diffusion, that
• 9:39
  potentially game -changing speed in AI text generation.
• 9:41
  Right, moving from sequential to diffusion -based
• 9:44
  text. Then we hit Bitcoin's massive rally, pushing
• 9:48
  past $110k for the first time. Driven by things
• 9:51
  like ETFs, institutional buying, and maybe a
• 9:54
  slightly better regulatory outlaw. And finally,
• 9:57
  we looked at the narratives around Sam Altman
• 9:59
  and OpenAI through those two new books, and how
• 10:02
  different generations are actually using tools
• 10:04
  like ChatGPT. Yeah, highlighting the growing
• 10:07
  influence and the evolving human -AI relationship.
• 10:10
  It really paints a picture of tech moving incredibly
• 10:13
  fast, affecting finance, work, even how we get
• 10:17
  advice. Absolutely. Constant change. Well, that's
• 10:20
  a quick look at some of the key tech news shaping
• 10:22
  our world today. We hope this deep dive has helped
• 10:25
  you connect the docs and understand what these
• 10:27
  developments mean for you. Yeah, hope it was
• 10:29
  useful. We are here to help here at Snarful,
• 10:32
  so reach out on our website if you have any more
• 10:33
  questions. That's a wrap on today's top stories.
• 10:40
  Remember, tech doesn't have to be complicated,
• 10:43
  and we're here to help connect the dots. If you've
• 10:45
  got questions or want to learn more, visit us
• 10:48
  at snarful .com. Thanks for listening, and we'll
• 10:51
  catch you next time.