text stringlengths 0 2.46k |
|---|
[4016.24 --> 4019.86] Are you for, I guess, then, people owning their own data? |
[4020.40 --> 4020.64] Oh, no. |
[4020.64 --> 4021.92] It seems that that's your... |
[4021.92 --> 4023.14] So, here's the thing. |
[4023.50 --> 4028.44] I'm absolutely for both from a Databricks perspective but also from an open source perspective. |
[4028.68 --> 4028.76] Right? |
[4028.94 --> 4029.10] Yeah. |
[4029.10 --> 4030.48] So, I'm an open source contributor. |
[4030.58 --> 4032.90] I contributed to Apache Spark and MLflow. |
[4033.22 --> 4035.24] And I'm also a maintainer for Delta Lake. |
[4035.58 --> 4035.74] Okay? |
[4036.10 --> 4037.28] And so, yeah. |
[4037.96 --> 4043.00] By definition, I'm always going to lean toward open source, which means you should own your data. |
[4043.12 --> 4044.30] Data should be a competitive advantage. |
[4044.84 --> 4047.74] Everything else should be open source, basically, for all intents and purposes. |
[4047.74 --> 4053.48] I'm even for things like differential privacy and privacy-preserving histograms to basically protect your data. |
[4053.66 --> 4056.86] And I can go on a diatribe on that, so let's not do that. |
[4057.28 --> 4064.82] But the context is, I'm not saying, though, these services like OpenAI or Bing or whatever else aren't worthwhile. |
[4065.44 --> 4065.92] They are. |
[4066.30 --> 4067.06] They're cheap. |
[4067.34 --> 4067.90] They're helpful. |
[4068.20 --> 4072.14] In fact, training other systems isn't necessarily a bad thing either. |
[4072.84 --> 4074.88] For me, it's not about don't do it. |
[4075.24 --> 4077.36] It's about knowing what you're doing. |
[4077.36 --> 4077.68] Right. |
[4077.86 --> 4078.20] That's it. |
[4078.30 --> 4078.38] Yeah. |
[4079.16 --> 4079.64] Transparency. |
[4079.72 --> 4079.92] Exactly. |
[4080.02 --> 4080.26] That's it. |
[4080.36 --> 4081.32] That's my con. |
[4081.90 --> 4086.32] If you want to use OpenAI within a database platform, we make it easy. |
[4086.52 --> 4091.24] We have a, for crying out loud, we add a SQL syntax directly so you can literally write Spark SQL, |
[4091.70 --> 4095.62] which basically is, at this point, is basically anti-SQL compliant. |
[4095.64 --> 4095.86] Right. |
[4095.86 --> 4102.52] You literally write SQL to go ahead and access your OpenAI to run an LL model directly against your data. |
[4102.86 --> 4104.92] So, literally, party hardy. |
[4105.12 --> 4105.46] Have fun. |
[4105.46 --> 4109.78] So, it's not, our attitude isn't so much like don't use one versus the other. |
[4109.78 --> 4112.70] Our attitude is very much, no, no, just know what you're doing. |
[4113.28 --> 4116.06] Understand when you're using something like a service. |
[4116.46 --> 4118.94] Understand when it makes sense for you to build your own model. |
[4118.94 --> 4124.32] And we also make it easy for you to build, maintain, train, infer against that model. |
[4124.60 --> 4124.94] That's it. |
[4125.60 --> 4127.94] So, I mentioned we have our transcripts as open source, right? |
[4128.02 --> 4128.42] Yeah. |
[4128.42 --> 4131.80] Everything we're saying here, when it hits the podcast, it's going to be transcribed into words. |
[4131.80 --> 4131.82] Exactly. |
[4132.42 --> 4138.40] How are ways we can use Dolly 2.0, this open model that you're talking about, this direction, |
[4138.94 --> 4143.60] how can we leverage these transcripts for our personal betterment as a podcast company? |
[4143.62 --> 4148.78] For example, as a podcast company, one of the first things, in fact, I'm actually already doing this technically for Delta Lake, okay? |
[4149.14 --> 4151.24] Is that we also have podcasts ourselves, okay? |
[4151.58 --> 4152.68] So, what are we doing, though? |
[4153.14 --> 4158.14] I'm spending time and effort to generate blogs based off of the podcast. |
[4158.56 --> 4158.88] Why? |
[4159.02 --> 4161.10] Because it's better for Google SEO search, right? |
[4161.76 --> 4164.32] It's not like I'm trying to just repeat the same thing. |
[4164.36 --> 4168.74] I'm just trying to summarize because, you know, we talked about barbecue in the beginning, right? |
[4168.74 --> 4169.52] We talked about coffee. |
[4169.52 --> 4174.74] We probably don't need all of those details inside the transcript of the podcast of our blog. |
[4175.14 --> 4178.54] You want people to go ahead and actually understand what they're talking about when it comes to Dolly, |
[4178.54 --> 4183.10] cool, we generate a blog based off of this conversation. |
[4183.36 --> 4185.64] It can summarize it, get to the key points. |
[4186.02 --> 4187.02] Boom, there you go. |
[4187.48 --> 4192.72] It simplifies the whole process so that way you're not spending exorbitant hours trying to figure out |
[4192.72 --> 4199.44] how to basically synthesize the key points out of our conversation right now, right? |
[4199.74 --> 4204.76] So, it's still time for you to review and look to make sure the model isn't giving you garbage. |
[4204.76 --> 4211.76] It's still time for a producer or for any other person who is knowledgeable in this field to validate the statements. |
[4211.86 --> 4214.06] Maybe I'm full of, you know, BS of all I know, right? |
[4214.30 --> 4216.10] And then so you get next to it and he's like, oh, yeah, yeah. |
[4216.36 --> 4216.82] I don't know. |
[4216.92 --> 4217.74] Denny's full of it. |
[4217.86 --> 4218.26] Forget it. |
[4218.50 --> 4221.46] It'd most likely be the Conical versus Flatbird Grinder. |
[4221.58 --> 4223.42] But, again, you know, that's a whole other story. |
[4223.48 --> 4224.42] The whole summary will just be added. |
[4224.42 --> 4226.32] I'm on your team Conical is me. |
[4227.16 --> 4227.82] I'm Conical. |
[4227.94 --> 4228.40] Team Conical. |
[4228.64 --> 4229.00] There you go. |
[4229.06 --> 4229.30] Perfect. |
[4229.42 --> 4229.50] See? |
[4229.86 --> 4234.12] But the context is that we can go ahead and actually use these systems to simplify. |
[4234.46 --> 4237.96] Would it be cheaper and easier if we just went ahead and did like ChatGB to do it? |
[4238.14 --> 4238.38] Yeah. |
[4238.90 --> 4239.40] Go for it. |
[4239.96 --> 4243.38] Would it be worthwhile to do it in your own Dolly model? |
[4243.52 --> 4243.86] Absolutely. |
[4244.02 --> 4247.50] Because you have your own style, right? |
[4247.70 --> 4247.84] Yeah. |
[4247.84 --> 4256.84] So if you have your own style, if it's building, if Dolly or any other open source model, again, I want to be very clear here, is going ahead and be trained against your transcripts. |
[4257.62 --> 4262.84] It will then be able to start writing blogs based off of your style, right? |
[4263.46 --> 4264.64] That's the cool thing about it. |
[4264.78 --> 4270.38] Is it cool to actually chain like that or is it better to go to a foundational model and then just our stuff? |
[4270.50 --> 4276.56] Or it'd be cooler to be like, well, start with Dolly because it has instructions and then add our style and then maybe add something else. |
[4276.56 --> 4278.70] I'm telling you my answer is all of the above because we don't know. |
[4278.70 --> 4279.36] Just whatever you want. |
[4279.40 --> 4279.54] No, no. |
[4279.54 --> 4279.86] We don't know. |
[4280.08 --> 4281.46] We don't know because that's the whole point. |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.