Latest / In the Loop AI / AGI is Back! Google Gemini 3.0 Crushed Our Expectations

Transcript

AGI is Back! Google Gemini 3.0 Crushed Our Expectations · In the Loop AI

0:00
Josh: If I had a crown in my hands, I would place it on the heads of Google because they have done it again.
0:04
Josh: They have the world's best AI model ever in history by a shockingly large margin.
0:09
Josh: Gemini 3.0 just got released.
0:11
Josh: It's available now to anybody in the world to go use it. And the benchmarks
0:15
Josh: are kind of blowing everyone's expectations out of the water, myself included.
0:18
Josh: And most importantly, it places another data point on the chart that shows we
0:22
Josh: are continuing to ascend up this exponential curve towards AGI.
0:26
Josh: And the roadmap is still intact and we are very quickly moving through it.
0:30
Josh: This, EJs, I was just, I was going through the benchmarks before recording this
0:32
Josh: and I, it's, it's shocking because we live in this world and yet somehow I'm
0:37
Josh: still continually blown away by the progress that's made by these models. So let's get into it.
0:41
Josh: Please walk everyone through, tell me, what did Gemini and the Google team just
0:45
Josh: release with this 3.0 update?
0:47
Ejaaz: People probably think we say the world's best model every week,
0:51
Ejaaz: but this time we really, really mean it.
0:53
Ejaaz: Like they have blown every single other model provider out the water.
0:58
Ejaaz: The things that this thing can do. Well, how about how about I just show you?
1:01
Ejaaz: How about I show you, Josh?
1:03
Josh: Please, let's see some examples.
1:04
Ejaaz: We have a thread here. And Sundar basically says, you can give Gemini 3 anything,
1:09
Ejaaz: images, PDFs, scribbles on a napkin, and it'll create whatever you like.
1:13
Ejaaz: For example, an image becomes a board game.
1:15
Ejaaz: A napkin sketch transforms into a full website.
1:18
Ejaaz: And a diagram could turn into an interactive lesson, right? So there's two examples
1:22
Ejaaz: I want to show you, Josh. I want to get your opinion on this.
1:25
Ejaaz: So number one, there's a short video of someone playing pickleball here and
1:30
Ejaaz: he or she rather uploads it into Gemini and says, hey, can you tell me how well
1:36
Ejaaz: I've done here and how I can improve my game?
1:38
Ejaaz: And it analyzes the entire video.
1:40
Ejaaz: It knows that she's wearing a knee brace. It analyzes her positions,
1:44
Ejaaz: telling her where she can move to better position herself to score the point.
1:48
Ejaaz: That's pretty nuts. But before I get your reaction to that, because Josh,
1:50
Ejaaz: I know you're an athlete. I know you're very competitive when it comes to these
1:53
Ejaaz: things. So this is a tool you could definitely use.
1:54
Ejaaz: The second thing is probably applicable to a lot of listeners on this show.
1:59
Ejaaz: They've embedded Gemini 3 into Google search and into new generative UI experiences.
2:05
Ejaaz: The way I would summarize this is it basically is very intuitive, Josh.
2:10
Ejaaz: It understands what you're asking for without you needing to really kind of explain yourself.
2:14
Ejaaz: The example they're showing on the video here is, can you explain the three body problem to me?
2:20
Ejaaz: And rather than just give you kind of like this simplistic text,
2:22
Ejaaz: which explains the concept, it decides to create a video diagram from scratch
2:27
Ejaaz: to show you a visual depiction of how this works.
2:30
Ejaaz: Right. Give me your reaction in order from one to two. So starting with the sports.
2:35
Josh: Okay, let's go. The first example.
2:36
Ejaaz: Yes, sir.
2:37
Josh: So this is really cool, the napkin example, where you can scribble something
2:40
Josh: down on a piece of paper, it'll generate it in the real world.
2:43
Josh: What all these examples are kind of showing me is what we always talk about
2:46
Josh: with Google, where it has this awareness of physics, reality,
2:51
Josh: and visuals and understanding what it's seeing.
2:53
Josh: And all three of these examples are leaning into that. So it leads me to believe
2:56
Josh: Gemini really is a multimodal first model, where it's meant to ingest,
3:00
Josh: meant to understand the world around us. This example of the chessboard and
3:04
Josh: the napkin is amazing because a lot of people oftentimes have sketches.
3:08
Josh: You just draw it down on paper and it intuitively understands it.
3:10
Josh: But the one that was most surprising to me is the video example.
3:13
Josh: Because as far as I'm concerned.
3:16
Josh: As far as I'm aware, there has never been a model that can ingest video and
3:20
Josh: understand the video that it sees. And if it does exist, I've never tried it before.
3:23
Josh: So the idea that you can, I mean, I play baseball growing up.
3:26
Josh: I, if I could take a video of myself swinging and get a corrective coach to
3:30
Josh: walk me through exactly what was wrong.
3:32
Josh: A lot of people play golf. I'm sure who are watching this.
3:34
Josh: If you could have a phone recording of you playing golf and it can actually
3:38
Josh: critique it and then critique me as if you are a tiger woods,
3:40
Josh: critique me as if you are whoever else is good that plays golf.
3:43
Josh: I don't know, Rory McIlroy, whatever they are, but like critique Critique me
3:45
Josh: as if you are an expert who is really good at golf and can give me some feedback
3:49
Josh: on how I can better my swing.
3:51
Josh: And what this offers in this one just narrow case example is now you have this
3:54
Josh: personal tutor that can do anything. If you're dancing, if you're doing anything
3:58
Josh: physical, if you're whatever it is, it can evaluate things for you.
4:01
Josh: Even if you have a video, a podcast, EJS, we uploaded to Gemini 3.0,
4:05
Josh: it could critique us. What did we do well? What did we not?
4:07
Josh: What did the visuals look like? How can we improve them? And that awareness
4:10
Josh: of video is like really cool.
4:13
Ejaaz: Yeah, I just want to say, I think the closest we got to this was with GPT,
4:18
Ejaaz: where you can upload an image, like what's under my car bonnet and say,
4:22
Ejaaz: hey, what's wrong? My car stopped working.
4:24
Ejaaz: And it can kind of like identify the point that you need to kind of like change,
4:28
Ejaaz: change the oil, blah, blah, blah.
4:29
Ejaaz: But that's just a static image. To go from that to live video and for it to
4:35
Ejaaz: analyze all the frames in that video and then give you a response on that is
4:38
Ejaaz: a massive leap upwards. We just haven't seen that anyway. So yeah, you're right.
4:42
Josh: It's amazing. And every example we go through, it kind of breaks the mold of
4:45
Josh: what I believe should be possible.
4:47
Josh: And I find that it's going to be difficult to use Gemini 3.0 because there are
4:50
Josh: so many possibilities now that have not existed previously. You kind of need
4:54
Josh: to relearn how to engage with AI because it's so capable.
4:57
Josh: And there's a fourth example here that I just want to touch on briefly,
5:00
Josh: which was also cool, is that it works just as well for the other things.
5:03
Josh: The example is a trip planning one where it starts to plan a trip and a vacation.
5:07
Josh: And it shows you a full list that is fully interactive of all the places broken up day by day.
5:12
Josh: And there's an option that you could just choose visual layout.
5:15
Josh: And you see on the screen here, it'll take every single day of your trip,
5:18
Josh: break it into images, section it out into this really nice visual grid.
5:22
Josh: So what I'm seeing the themes here are, okay, real world understanding,
5:25
Josh: video first, and really nice presentation, which I think a lot of models sometimes struggle on.
5:31
Josh: Demo's out of control. I'm excited to use it. Everyone else can use it now.
5:35
Josh: It's live. Now I want to get to benchmarks, EJS, because this is where things
5:39
Josh: get kind of crazy, where we could actually compare one to another and see exactly
5:43
Josh: how impressive this is relative to everybody else. So please, we have the card here.
5:47
Josh: Walk us through what we're seeing in this model card and all the specs that we need to know.
5:50
Ejaaz: As you guys probably know by now, benchmarks is typically how we evaluate any
5:54
Ejaaz: typical AI model against each other.
5:57
Ejaaz: And they're measured against a range of different benchmarks.
5:59
Ejaaz: A benchmark can be considered as sort of like a test.
6:01
Ejaaz: Now, right at the top, you've got humanity's last exam. This is by default,
6:06
Ejaaz: the hardest exam that an AI model is tested against.
6:10
Ejaaz: And it's kind of like an academic reasoning test with no tools accessible to it.
6:14
Ejaaz: It scored a very impressive 37.5%, which is more than I think is about a 15%
6:21
Ejaaz: increase from its previous model.
6:23
Ejaaz: Very, very impressive. But what really blew my mind was the second stat listed
6:28
Ejaaz: here, which is the ARK AGI 2 benchmark.
6:31
Ejaaz: Josh, when I say this 2x'd the previous state-of-the-art model, I absolutely mean it.
6:38
Ejaaz: In fact, let me just show you this chart here.
6:41
Ejaaz: Now, you may notice a couple of bustly specs here.
6:45
Ejaaz: GPT-5 Pro, Grok 4 Thinking. And then can you see that outlier right at the top
6:51
Ejaaz: right. Do you see that, Josh?
6:53
Josh: That's insane. The two outliers.
6:55
Ejaaz: The two outliers. So these are the Gemini 3 Pro and the Gemini 3 deep thinking model.
7:02
Ejaaz: Deep thinking being like, you know, a large number that can like kind of give
7:04
Ejaaz: you a more research response. They are a...
7:07
Ejaaz: Stand out from every single other model. And the reason why this is so crazy,
7:12
Ejaaz: well, there's a few reasons.
7:13
Ejaaz: Number one, all the other model progressions, as you can see over time,
7:18
Ejaaz: has been kind of impressive, but kind of small.
7:23
Ejaaz: Like they've been a good jump. It's been impressive, but it hasn't been as impressive
7:28
Ejaaz: to be like, oh, you know, another model provider couldn't catch up.
7:31
Ejaaz: These results from Google literally put it miles ahead of every other model.
7:36
Ejaaz: So when I sat at this chart, I think, wow, Google probably has the lead for
7:40
Ejaaz: another six months and in six months time, they're going to have a more impressive model by then.
7:46
Ejaaz: So at this point, I'm kind of thinking, can anyone catch up to Google?
7:49
Ejaaz: Josh, do you have any reactions to this benchmark?
7:51
Josh: This is the chart that like the first thing I said to myself when I saw this
7:55
Josh: is like, oh, my God, there is no wall there.
7:57
Josh: We are not going to stop scaling. The scaling will still apply because these
8:01
Josh: two new data points that we have blow everything else out of the water.
8:03
Josh: And this is how exponential growth happens.
8:05
Josh: It seems like a really small cluster down there in the bottom,
8:08
Josh: but the reality is that was the top just a couple hours ago.
8:12
Josh: And Gemini kind of refactored this entire chart to make it seem like it's so
8:17
Josh: small because the progress is so high.
8:18
Josh: And although Gemini 3.0 thinking is seemingly the most impressive,
8:23
Josh: the really anomaly chart is the Gemini 3.0 Pro, which is basically a vertical
8:27
Josh: line up from these other models, where the score is higher, but the cost is
8:32
Josh: actually slightly lower.
8:33
Josh: And if you connect a dot between these averages, you start to see literal vertical
8:37
Josh: line in terms of improvement and acceleration in these models.
8:40
Josh: And that to me shows that there is no scaling wall that we're hitting.
8:45
Josh: Like we can continue to scale resources, energy, compute, and we could continue
8:48
Josh: along this path towards AGI in a world where some people were saying,
8:51
Josh: we don't know if it continues. The answer to me is very clearly, it continues.
8:55
Josh: This is a step much closer to AGI.
8:58
Josh: And again, that real world understanding makes it feel much closer to AGI than it ever has before.
9:04
Josh: Because now it really like intuitively understands the world through video,
9:08
Josh: through photo, through audio, through basically every sensory input we have
9:11
Josh: outside of what taste and feel.
9:13
Josh: So this to me, I saw this chart. I was like, oh, my God, Gemini,
9:17
Josh: you really outdid yourselves.
9:18
Ejaaz: I'm just going to be honest. I think over the last couple of months,
9:22
Ejaaz: I've been getting a little bored with the models that have been released by other model providers.
9:27
Ejaaz: And it led me to think that we're not going to make many breakthroughs until
9:32
Ejaaz: some model provider figures out a new, unique way to train their model.
9:37
Ejaaz: Gemini or Google has convinced me otherwise with this release.
9:41
Ejaaz: But I know you guys are probably like fed up with listening to us hop on about benchmarks.
9:45
Ejaaz: So how about I materialize that for you in a much more easy to understand way, right?
9:50
Ejaaz: So here are the four big takeaways that you need to learn about Gemini 3.
9:55
Ejaaz: Number one, for its intelligence, for the intelligence that you're getting,
10:00
Ejaaz: it is not that super expensive.
10:03
Ejaaz: Google trained this from scratch, as this tweet says, using their own GPU infrastructure.
10:07
Ejaaz: And it used this kind of like layout called a mixture of experts,
10:11
Ejaaz: which basically means that whenever you prompt the model, it's not going to use the entire model.
10:16
Ejaaz: So it actually ends up being cheaper than what it could eventually become.
10:19
Ejaaz: One million token context input, 64K token output. We'll get to the costs in
10:24
Ejaaz: a bit of a second, But the point that I'm making here is that it's not as expensive
10:27
Ejaaz: as you would expect for the intelligence that you're getting.
10:29
Ejaaz: Now, if you compared Gemini 3 to GPT 5.1 from OpenAI, on a relative basis,
10:36
Ejaaz: it is more expensive. But for the jump in intelligence that you're getting,
10:39
Ejaaz: it's way better. So it's, in my opinion, worth it.
10:42
Ejaaz: Number two, when it comes to computer use, so that means letting the AI model
10:47
Ejaaz: control your computer and do tasks for you whilst you go do something else.
10:52
Ejaaz: It is state of the art. It is the best here. They measured it against a benchmark
10:56
Ejaaz: called ScreenSpot Pro, which is a benchmark which kind of like analyzes its
11:00
Ejaaz: ability to understand images and visuals on a desktop. It just absolutely crushes it.
11:06
Ejaaz: Number three, it is the best AI for math by far.
11:09
Ejaaz: So again, the point I'm making here or the theme that we're seeing here is it's
11:13
Ejaaz: not just good at one thing, it's good at many things, which makes it the best
11:17
Ejaaz: generalist AI model in the world right now, by far.
11:21
Ejaaz: And the final thing, Josh, and this is where it might slip up.
11:24
Ejaaz: I'm curious to get your take on this.
11:26
Ejaaz: It is insanely good at coding, but we don't quite know if it is the best at coding yet.
11:32
Ejaaz: What I mean by that is it completely crushed everyone else on one coding benchmark,
11:37
Ejaaz: but the coding benchmark that matters, which is the software engineering,
11:40
Ejaaz: SWE, it didn't do as well as its competitor, Claude 4.5 from Anthropic.
11:45
Ejaaz: So those are the four main takeaways.
11:47
Josh: I would much prefer a model that understands the world than understands how to code.
11:51
Josh: And I think we're starting to see these subset niches where if Anthropic has
11:55
Josh: the best coding model, that's great.
11:56
Josh: Let them focus on code. Let them narrowly make that the best model.
12:00
Josh: Let Google handle everything else. And I think that's what Gemini is focusing
12:03
Josh: on. So the code thing doesn't really bother me because I don't care to use Gemini for code.
12:07
Josh: I'm happy to be in Claude Camp for code and then use Gemini for everything else.
12:11
Josh: And then one of the points earlier that you mentioned on the pricing is.
12:14
Josh: I find it a little interesting because it's a little bit more than just a little bit more expensive.
12:18
Josh: The pricing, I was looking through it and it's for over 200,000 tokens.
12:22
Josh: They're charging $4 for inputs and $18 for outputs.
12:27
Josh: Now, relative to GPT 5.1, which just got released, they're charging for a million
12:32
Josh: tokens, $1.25 in, $10 out.
12:35
Josh: So you're talking about, what is that? that's about $20 versus $1.25 on inputs.
12:42
Josh: And that is a fairly significant margin that you're paying for this quality.
12:45
Josh: So we're starting to see the trade-offs happening on that Pareto curve that
12:48
Josh: we talked about in a few episodes earlier, where there are trade-offs coming
12:52
Josh: in terms of cost and quality.
12:53
Josh: And it's clear that while OpenAI may have optimized for cost,
12:57
Josh: Google is kind of optimizing for a little further up the cost curve in exchange for super high quality.
13:02
Josh: And it seems like this is kind of a balanced data point for now because unless
13:06
Josh: you are using this via API and you're requiring a ton of tokens,
13:09
Josh: a $20 a month Google membership will get you all of the use that you need. And that is just fine.
13:14
Josh: So in terms of a usability perspective, I think that's okay.
13:18
Josh: But it's just an interesting thing to know is that this is a better model.
13:21
Josh: It is also more expensive.
13:22
Josh: And that is a trade-off that was made. And in the case that OpenAI decides to
13:25
Josh: make this trade-off with GPT-6 or Grok decides to make this with Grok-5,
13:30
Josh: or Grok 6. I'm losing track of all these models now. I think we're going to
13:34
Josh: start to see the dynamic shift in terms of that Pareto curve and what model
13:37
Josh: architects decide to remove and add.
13:39
Josh: And in this case, it looks like Google added quality, but they also did add
13:42
Josh: quite a significant cost increase.
13:44
Ejaaz: I personally don't think it matters. I think it's a nothing burger.
13:46
Ejaaz: I think that if Google wanted to make it affordable for everyone,
13:50
Ejaaz: including the developers that want to get API access, that think it might be
13:52
Ejaaz: too expensive, they could subsidize it.
13:54
Ejaaz: They are a cash flow giant. They have enough money to do that.
13:57
Ejaaz: OpenAI has been doing that for so long now that it doesn't even matter.
14:00
Ejaaz: I don't see any reason why Google couldn't do that.
14:02
Ejaaz: The other reason is Google just released their latest TPU, which is their GPU
14:08
Ejaaz: that they use to train their models and inference their models.
14:11
Ejaaz: And typically with every generation, we get a much cheaper cost of inference.
14:15
Ejaaz: I think by the time that they release their next generation model,
14:18
Ejaaz: which might be, you know, Gemini 3.1, we're going to see a considerable reduction
14:24
Ejaaz: in the cost for using Gemini 3 Pro and Gemini Pro deep research.
14:29
Ejaaz: So I'm not too worried about that. I think it's kind of like a short-term problem
14:32
Ejaaz: and not a long-term problem.
14:34
Ejaaz: But kind of speaking of TPUs, I just want to take a moment to really kind of
14:39
Ejaaz: belay the point that using their own TPUs to train a state-of-the-art model
14:44
Ejaaz: that is 2x better than the previous state-of-the-art model and probably puts
14:48
Ejaaz: them in a six-month lead after Google started off on the back foot,
14:53
Ejaaz: creating probably the worst model I've ever seen and
14:56
Ejaaz: changing that all around in what's it under two years is
14:59
Ejaaz: nothing short of insanity tpus is
15:02
Ejaaz: uh google's kind of version of the gpu gpu is
15:05
Ejaaz: kind of like what nvidia controls the monopoly over this is the hardware that
15:09
Ejaaz: you use to train your ai and inference your ai the uh unique part here is that
15:13
Ejaaz: google's never used an nvidia gpu uh or in any considerable way to train their
15:19
Ejaaz: models they've always trained it in-house and that's such a difficult and tricky
15:23
Ejaaz: thing to do because designing and building these
15:25
Ejaaz: TPUs at scale, these GPUs at scale is a super hard and complex thing.
15:29
Ejaaz: You need so much talent, you need so much expertise and insight to be able to do that.
15:33
Ejaaz: The unique thing about Google's TPUs, well, there's two main takeaways.
15:36
Ejaaz: Number one, it's cheaper to train the same amount of intelligence that an NVIDIA
15:41
Ejaaz: GPU is. So it's more cost efficient.
15:43
Ejaaz: And the second way is, and this is their secret sauce, you can stack those TPUs
15:47
Ejaaz: on top of each other in a really scalable way that you can start training really, really large models.
15:53
Ejaaz: If you wanted to train the same size model with nvidia
15:57
Ejaaz: gpus it would cost way more and it would take way
16:00
Ejaaz: longer so google made a really risky and big bet about a decade ago saying we're
16:05
Ejaaz: going to build our infrastructure in-house and we're not going to rely on nvidia
16:07
Ejaaz: and we're going to benefit from the full stack experience and this model is
16:11
Ejaaz: a prime example of that bet paying off so i just want to call them out like
16:15
Ejaaz: it's not like google has gotten lucky here they've been planning it for a while now The
16:19
Josh: Interesting thing to me is that this is the first number one model in the world
16:24
Josh: built on something other than an NVIDIA GPU.
16:27
Josh: And that's fairly significant because every company in the world is trying,
16:31
Josh: but this is proof that it's actually possible.
16:32
Josh: And I think when we talk about Tesla and AI5 and the XAI team,
16:37
Josh: when we talk about OpenAI working with whoever they're working with to build
16:41
Josh: their own in-house GPUs, I think this sets a precedent that it is possible.
16:45
Josh: And I suspect that will result in more companies putting their foot on the gas.
16:49
Josh: When it comes to kind of destructing part of NVIDIA's monopoly that it holds
16:53
Josh: over GPUs. So that to me is the interesting takeaway of this.
16:56
Josh: And hearing that it was fully done, trained on these TPUs, that's very high
17:00
Josh: signal to me saying, okay, there is an architecture chips happening.
17:03
Josh: There is a real benefit to vertical integration if you could figure out manufacturing
17:06
Josh: these compute units at scale.
17:08
Josh: And now the race is on for everyone to do this.
17:12
Josh: Because again, using the Apple example, the M series chips, unbelievable,
17:15
Josh: and they unlocked the best computers in the world.
17:16
Josh: And if companies can really start to refine this vertical integration of their
17:20
Josh: own chips, you're going to see that exponential curve go vertical times 10.
17:24
Josh: Like it is going to, I suspect that is very...
17:28
Josh: Obviously now how we reach AGI faster than people previously thought,
17:32
Josh: because the efficiency improvements from those vertical integrations,
17:35
Josh: once they're able to manufacture these at scale, are going to be unbelievable.
17:39
Josh: And I'm so excited for that to happen in the near future. Google has a big head
17:43
Josh: start, but let me tell you, the other companies are not far behind.
17:46
Ejaaz: Well, let me introduce you to another big advantage of being the big dog, Google.
17:52
Ejaaz: You thought you were going to come on to this episode and listen to us hopping
17:55
Ejaaz: on about a generalized model? No.
17:57
Ejaaz: You're forgetting Google has many other products in their arsenal,
18:00
Ejaaz: and you're forgetting that they can plug in their new state-of-the-art model into all of them.
18:05
Ejaaz: So Google not only today announced Gemini 3, but they also announced a different product.
18:10
Ejaaz: It's called Google Anti-Gravity, which is basically a new software environment
18:15
Ejaaz: for you to code up AI agents, except this time these AI agents are going to
18:20
Ejaaz: be super, super smart because they get plugged in with Gemini 3.
18:24
Ejaaz: Now, if you remember earlier, I mentioned that one of the cool benchmarks that
18:27
Ejaaz: this new model sets is in computer use, which means that it can control your
18:31
Ejaaz: computer, it can do things autonomously for you.
18:34
Ejaaz: Now, typically, the reason why we haven't really spoken about that on the show
18:36
Ejaaz: is that they've been kind of lame.
18:38
Ejaaz: Like they can book you a dinner reservation and do different kinds of stuff.
18:42
Ejaaz: With this model, it's way more intuitive. It can do way more intelligent tasks
18:46
Ejaaz: and it can take a lot more complex work off of your hands such that the value
18:50
Ejaaz: that it produces to you over like the eight hours that you take to sleep overnight
18:54
Ejaaz: would be considerable for you to actually be serious to use in your enterprise,
18:58
Ejaaz: in your business, or just at your at-home lifestyle, right?
19:01
Ejaaz: So the point I want to make around here is Google's moat is not just its intelligence
19:07
Ejaaz: or ability to create new models. It's not its TPUs.
19:10
Ejaaz: It's its distribution. It's the entire product suite that it has that regular
19:15
Ejaaz: users like you and I that use Gmail, that use Google Suite can now kind of benefit
19:19
Ejaaz: from simply by plugging in that model.
19:22
Ejaaz: And I think like products like this, anti-gravity, I bet you,
19:25
Ejaaz: Josh, we're going to see a slew of new Google product releases over the next
19:28
Ejaaz: couple of weeks simply because they created this model.
19:32
Josh: I hope so. I guess the contrarian take is like, okay, how many people are actually
19:36
Josh: going to want to use them?
19:37
Josh: We just spoke about how Claude is the superior code model. Everyone loves Cursor.
19:42
Josh: No one really uses the mobile applications of these. A lot of people are engaging
19:46
Josh: with AI on their phone. So maybe it works for the right type of person.
19:49
Josh: But Google still does have that product problem where they kind of have a tough time.
19:54
Josh: They have the amazing intelligence. They just have a tough time using it.
19:56
Josh: I mean, I don't have the Gemini app on my phone. I mostly use Grok and ChatGPT.
20:00
Josh: And there is this bar that they
20:01
Josh: still need to cross that I think they're trying with Google AI Studio.
20:05
Josh: And we had Logan Kilpatrick on, who was the head of that, to talk about it when Nano Banana came out.
20:08
Josh: But there is still a bit of a long shot for them to get good at products to actually develop this.
20:14
Josh: But what we saw this week is that there was a resounding, overwhelming amount
20:17
Josh: of support, to your point, you guys, where the market just believes in Google.
20:20
Josh: And in a week where all of the stocks, all of the Mag7 was down,
20:24
Josh: Google was the one anomaly. Google was up this week.
20:26
Josh: And I think it's because the market is starting to realize, one,
20:29
Josh: vertical integration through these TPUs is a huge deal.
20:33
Josh: Two, Google has an existing business that is not reliant on AI.
20:36
Josh: And sure, AI places a huge hand on that scale, but it is not everything.
20:41
Josh: And they are cashflow positive in the absence of AI. So all of this innovation
20:45
Josh: that they're doing is really just applying later fluid on top of an already
20:48
Josh: great business. And the market's starting to evaluate that properly.
20:51
Josh: So Google is positioned very strongly.
20:53
Josh: They have very high intelligence. Gemini 3 rocks.
20:56
Josh: And I mean, again, we continue on the bull train for Google.
20:59
Josh: I am a believer. I am a supporter.
21:01
Josh: I am stoked that they have the crown. I assumed it was only a matter of time.
21:04
Josh: And now the question is, who's next?
21:06
Josh: Who is the next competitor? Who's going to set the next plot on that chart and
21:10
Josh: set the vertical trajectory on the exponential curve we're on?
21:14
Josh: Do you have any guesses who you think it's going to be?
21:16
Ejaaz: Yeah, well, I don't because I don't think it's going to be anyone for a while.
21:20
Ejaaz: I said this earlier in the show and I'm going to say it again. I think there's going
21:23
Ejaaz: to be a six month period now where either the other model providers don't release
21:29
Ejaaz: the model because it's not as good as Google's or they just kind of release
21:33
Ejaaz: these kind of mediocre kind of like consumer products that kind of maybe benefit
21:38
Ejaaz: certain consumers in one way or another,
21:41
Ejaaz: but doesn't really kind of break the generalized model
21:44
Ejaaz: standard that Google has just set.
21:47
Ejaaz: Just a last point on the kind of Google bull case thesis, they may not just
21:51
Ejaaz: play in the same ring as cursor does.
21:53
Ejaaz: Like I was critiquing Microsoft on another episode, Josh, do you remember?
21:57
Ejaaz: And then I got off that episode and I was just like, Microsoft like dominates
22:02
Ejaaz: the enterprise environment.
22:04
Ejaaz: All the boomer companies and institutions love Microsoft and they have all their data and memory.
22:09
Ejaaz: And just because you and I don't use it or just I'll speak for myself,
22:12
Ejaaz: just because I don't use it and I think it's doesn't mean that they're not absolutely crushing.
22:15
Ejaaz: Google just came off a hundred billion dollar quarter of revenue.
22:19
Ejaaz: That's like the highest they've ever had. So I don't want to be too hasty to
22:23
Ejaaz: say that like Google's not going to make it because they can't make a sick consumer
22:26
Ejaaz: product like OpenAI can maybe.
22:28
Ejaaz: I just think they're maybe playing in different fields.
22:31
Ejaaz: But to the point around like I don't think anyone else is going to catch up.
22:34
Ejaaz: Look at these comments. I want to show you two comments all right one is from sam altman
22:38
Ejaaz: He goes, congrats to Google and Gemini 3. This looks like a great model.
22:42
Ejaaz: The other is from the almighty being, Elon Musk, saying, I can't wait to try this out.
22:47
Ejaaz: And this is just one of a series of tweets that he's been putting out this week
22:49
Ejaaz: saying, can you guys just drop Gemini 3? Because I need to see how good this thing is.
22:53
Ejaaz: And the reason why I bring up these two people is both Sam Altman and Elon Musk
22:57
Ejaaz: have released new versions of their models, GPT and Grok respectively,
23:02
Ejaaz: but it's been the 0.1 upgrade.
23:04
Ejaaz: It's GPT 5.1. It is Grok 4.1. and they are almost identical updates.
23:10
Ejaaz: You want to know what the biggest and coolest thing about their model updates were?
23:13
Ejaaz: Personality traits, which don't get me wrong, is cool. Like I would like my
23:16
Ejaaz: model to kind of respond in a very intuitive manner and get me,
23:19
Ejaaz: but it's nowhere near the state of the art standard that we've just seen broken by Gemini.
23:23
Ejaaz: So the point I'm making is, I think these two companies might have run out of fuel for the near term.
23:28
Josh: Grok is going to be next. You think? They're the next one. By the end of Q1,
23:32
Josh: Grok will have the crown.
23:33
Josh: Why? And I assume by a fairly large margin. But I assume it will be a different type of crown.
23:39
Josh: And this is where I'm really excited to see how these models progress.
23:42
Josh: We spoke a little bit earlier about how Cursor is kind of the coding model.
23:46
Josh: Google has a very deep understanding of the real world and physics and video
23:50
Josh: and understanding how that works.
23:51
Josh: Grok and the XAI team are very focused on the pursuit of truth and information.
23:56
Josh: And I think that's kind of the alley that we'll see them going down.
23:59
Josh: So they have the real-time data with X. They have the pursuit of truth.
24:03
Josh: And where Google and OpenAI and all these other companies are trained on an
24:06
Josh: existing data set, the XAI team and the Grok team are developing an entirely
24:10
Josh: new synthetic data set that is maximally truth-seeking.
24:15
Josh: And we saw that early version with Grokopedia that should provide the most accurate
24:19
Josh: and I guess thoughtful information. It should be the best of thinking because
24:23
Josh: it's the closest to source truth. So while I think,
24:26
Josh: Gemini will probably be better at physics and video and understanding the real
24:30
Josh: world for quite some time.
24:31
Josh: I suspect Grok will be really good at just communicating via text.
24:36
Josh: If text is a modality in which we interface from, Grok should be really good.
24:40
Josh: And again, the rate of acceleration, Grok has been around for the least amount
24:42
Josh: of time. They're accelerating the fastest.
24:44
Josh: And I'm very, very, very excited for a Grok 6, Grok 5, whatever we're at announcement,
24:49
Josh: hopefully early next year. So that's the predictions.
24:52
Josh: That's the episode. That's Gemini 3.1. It is an unbelievable new model.
24:56
Josh: Everyone could try it out. So here's how you try it out.
24:58
Josh: I believe you need to be a Google premium plus subscriber, whatever it's called. It's $20 a month.
25:04
Josh: You can go on the Gemini website and it's just a text box and you can play around
25:07
Josh: with it. They also have a mobile application.
25:09
Josh: It's very easy to download on your phone, play around with it.
25:12
Josh: I'd love to see examples of cool things because I think one of the problems
25:15
Josh: for me and one of the things I'd love help with from anyone who's listening
25:17
Josh: is how do you use this thing to test it?
25:20
Josh: What do I ask it? And how are you interfacing with it to get the maximum amount of results from it?
25:25
Josh: Because intuitively, I would never think to record myself and ask for feedback,
25:28
Josh: but that's a new possibility.
25:29
Josh: So I guess the challenge to anyone who's listening is figuring out how to get
25:33
Josh: the most out of these new models as these new features get released.
25:36
Josh: And Gemini 3 has just opened up the gates to a gazillion new use cases.
25:41
Ejaaz: Yeah, I mean, this is a super cool release for Google.
25:46
Ejaaz: And weirdly enough, it's not the only release over the last week.
25:49
Ejaaz: I mean, I've got a list pulled up here. They've released new Android iOS updates.
25:53
Ejaaz: They've got a new search AI mode. They've released anti-gravity that we mentioned earlier.
25:57
Ejaaz: We've got CIMA2 research, which we demoed on a previous episode.
26:00
Ejaaz: You should definitely go check that out.
26:02
Ejaaz: I mean, they are just not stopping and they're a force to reckon with.
26:07
Ejaaz: And kind of similar to them, Josh, just to kind of round this episode out and
26:11
Ejaaz: thank you guys for listening.
26:13
Ejaaz: We are here in Argentina, in
26:15
Ejaaz: Buenos Aires. We are kind of meeting some of the fans that are out here.
26:18
Ejaaz: And we spoke to one just this afternoon, Josh. And you know what he said to me? Have a guess.
26:23
Josh: What's that?
26:24
Ejaaz: He said, your podcast Limitless is like the state-of-the-art AI podcast.
26:30
Ejaaz: In fact, it is 2X better than any other AI podcast that I've ever heard. And you know what?
26:35
Ejaaz: That sounds very similar to
26:36
Ejaaz: the Gemini 3. So you could potentially call us the Gemini 3 of AI podcast.
26:41
Ejaaz: And so if you're a listener to this, if you are a non-subscriber on our YouTube,
26:46
Ejaaz: you should probably click that subscriber button.
26:48
Ejaaz: You should probably click that notification button because guess what?
26:51
Ejaaz: We've got more episodes coming this week. And guess what? the five star ratings help us out massively.
26:56
Ejaaz: So if you enjoyed this episode and if you want to hear more episodes of this
26:59
Ejaaz: nature and of cutting edge news in AI, you should give us a follow and we will see