text stringlengths 0 2.46k |
|---|
[3780.74 --> 3781.84] But this is actually me being positive. |
[3781.84 --> 3791.90] They actually have introduced concepts saying you can pay more to train and that they'll never actually use your data. |
[3792.38 --> 3795.80] But I don't remember the cost, but it is definitely paying more. |
[3796.18 --> 3796.38] Yeah. |
[3796.60 --> 3796.76] Yeah. |
[3797.68 --> 3800.64] Well, it's not as valuable to them, so it makes sense as a transaction. |
[3800.78 --> 3801.02] Exactly. |
[3801.20 --> 3802.78] So, that becomes more of a transaction that way. |
[3802.84 --> 3803.04] Exactly. |
[3803.82 --> 3806.86] So, have you seen the Googler's leaked memo about we have no moat? |
[3807.34 --> 3807.56] Yes. |
[3807.56 --> 3808.06] Because isn't this like… |
[3808.06 --> 3809.82] Everybody talks about that memo. |
[3809.82 --> 3812.90] And what's interesting about that whole concept is that… |
[3812.90 --> 3816.50] I know it sounds sideways, but I was about to actually give you another context. |
[3816.70 --> 3817.92] And this is actually, again, my con over. |
[3818.04 --> 3818.60] I want to give credit. |
[3818.78 --> 3820.22] Attributions of the guy who actually said it. |
[3821.12 --> 3824.76] What's really interesting about this whole thing, when they talk about moat, they talk about everything else, |
[3824.80 --> 3829.46] is that more fundamentally, we could have done this two years ago. |
[3830.46 --> 3837.74] We could have taken this concept of basically saying small amount of data, foundational model, fine-tune it, |
[3837.74 --> 3839.66] and actually have good results. |
[3839.82 --> 3843.88] So, all of us were focusing on, I need a bigger model. |
[3844.12 --> 3845.26] I need to dump more data. |
[3845.58 --> 3850.34] I need to scrape the entire freaking internet and chuck it all into the gigantic model. |
[3850.60 --> 3857.00] Spend tens of millions of dollars, warp every single GPU until Azure basically melts in order to go ahead and train this thing. |
[3857.10 --> 3858.06] Until the heat death of the universe. |
[3858.08 --> 3858.58] Right, exactly. |
[3859.02 --> 3865.50] And then meanwhile, it's like, or we literally could have taken a foundational model that was okay to good, |
[3867.10 --> 3870.06] $100, and bam, we get something good. |
[3870.14 --> 3870.28] Yeah. |
[3870.28 --> 3870.32] Yeah. |
[3870.66 --> 3874.70] So, when they talk about, like, there's no moat and all this other stuff between open source and not, |
[3875.18 --> 3878.34] literally my attitude toward this whole thing is like, no, just step backwards for a second. |
[3878.74 --> 3878.94] Okay? |
[3879.34 --> 3881.00] The reality is we could have done this. |
[3881.26 --> 3887.40] We all got attracted to the idea, the shiny thing of, ooh, bigger, more, bigger, more, larger, more. |
[3887.40 --> 3888.70] That's all we got attracted to. |
[3889.36 --> 3892.34] And so, in the end, I'm going, I don't care. |
[3893.60 --> 3898.68] Like, these companies, the ones that, quote unquote, are trying to build a moat around themselves, |
[3899.22 --> 3905.38] what they're doing, they're trying to make sure that they have a service in which you will give them your data, |
[3905.78 --> 3909.46] and then by definition, you will give away your competitive advantage. |
[3909.58 --> 3909.76] Right. |
[3910.00 --> 3910.52] Simple as that. |
[3910.52 --> 3917.56] For the folks that don't want to do that, which I think is the vast majority, then my attitude is, like, quite simple. |
[3917.74 --> 3920.02] Then don't do that and build your own model. |
[3920.16 --> 3922.28] Now, how about if I'm the general consumer? |
[3922.48 --> 3926.98] I just want to pump out a good blog template for me to work with. |
[3927.52 --> 3927.76] Yeah. |
[3928.36 --> 3928.76] Absolutely. |
[3929.68 --> 3930.22] Why not? |
[3930.38 --> 3930.48] Yeah. |
[3930.80 --> 3933.60] Seriously, I'm not trying to say these services aren't worthwhile. |
[3933.86 --> 3934.46] Quite the opposite. |
[3935.00 --> 3935.60] ChatDB is fine. |
[3935.62 --> 3936.10] Very valuable. |
[3936.10 --> 3936.46] Oh, yeah. |
[3936.54 --> 3937.62] It's extremely valuable. |
[3937.62 --> 3941.90] In fact, I've already had it pumping out code for me just for shits and giggles. |
[3942.18 --> 3942.28] Yeah. |
[3942.64 --> 3943.68] So my Rust is- |
[3943.68 --> 3945.54] It's going to pump out some slides for you here soon for tomorrow. |
[3945.56 --> 3946.20] That's a good idea. |
[3946.20 --> 3947.20] I should test out that. |
[3947.44 --> 3947.72] Yeah, yeah. |
[3948.02 --> 3948.96] Take that 30 minutes. |
[3949.04 --> 3949.76] Turn it into 12. |
[3949.82 --> 3950.16] Oh, yeah. |
[3950.16 --> 3950.80] That'd be perfect. |
[3950.88 --> 3951.08] Yeah, yeah. |
[3951.18 --> 3952.02] But see, you get my drift. |
[3952.38 --> 3952.94] Yeah, totally. |
[3953.14 --> 3953.26] Yeah. |
[3953.32 --> 3955.10] So my Rust code is Rusty. |
[3956.22 --> 3960.86] And so basically, I was using ChatGBD to basically pump out a bunch of Rust code for me. |
[3960.88 --> 3962.68] I'm like, hey, this is a great boilerplate. |
[3963.06 --> 3965.44] Now I've got something to work with, and boom, now I can start writing again. |
[3965.54 --> 3965.74] Right. |
[3966.04 --> 3966.14] Yeah. |
[3966.14 --> 3969.14] So what is Databricks' play in this chess game? |
[3969.26 --> 3970.34] Like, what's your guys' angle? |
[3970.58 --> 3971.56] Our angle's quite simple. |
[3971.96 --> 3974.48] You've got a ton of data. |
[3974.90 --> 3977.54] You need to ETL and process it in the first place. |
[3977.96 --> 3985.78] Then you need to have a platform to run machine learning or data science or AI or whatever freaking wording you want to use. |
[3985.78 --> 3986.14] Okay? |
[3986.38 --> 3996.76] Whether it's LLMs today, deep learning yesterday or tomorrow, image optical resolutions, object recognition. |
[3996.94 --> 3997.66] I don't care. |
[3997.90 --> 3998.06] Okay? |
[3998.20 --> 4000.30] The point is that you have a ton of data. |
[4001.04 --> 4002.58] You need to be able to process it. |
[4002.76 --> 4007.18] You need to be able to access every single open source system or service. |
[4008.68 --> 4009.82] Databricks' play is quite simple. |
[4009.82 --> 4011.40] We just make it easy for you to do any of it. |
[4011.74 --> 4011.92] Yeah. |
[4012.06 --> 4012.44] That's it. |
[4012.66 --> 4013.84] That's our only play. |
[4013.96 --> 4014.72] Let's make it easy. |
[4015.16 --> 4015.28] Yeah. |
[4015.42 --> 4015.56] Yeah. |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.