31–40 of 78 posts
It seems AGI is the perpetual energy device or philosopher's stone of our era.
Enough so that I'd consider starting a hedge fund to short all the AGI companies :)
Ugh. Maybe I'm in a different world by now, but I dislike such statements on multiple levels. > This typically comes with assertions such as "data is the new oil" and "once we have our dataset and models the Big Tech shops will have no choice but to acquire us". Maybe it's me, but I dislike the attitude to work to be acquired. Interestingly, this is a rift I see quite a bit if I interview more development oriented gu…
Earlier quoted context omitted.
Yeah, people say that, but it's hard to see what selling shovels would amount to. I mean, hardware/cloud is commoditized, ML software is free. Maybe Uber for annotators /s (read, Mechanical Turk). (Of course, there is the ever more popular route of making services for enterprises wanting to outsource basic stuff.)
Realistically the "shovels" are the compute services. And the shovel makers are really doing well. What you don't really see is the ML startups making billions yet. What you do however see is companies like Microsoft with their Azure product and Amazon with AWS Compute making bank off the tech startups doing ML. I really think the people who are cashing in the most with the ML craze are the cloud/compute providers. I…
And Nvidia
Because there is a real moat with data ownership and pipelines. If you want to do any analysis you quickly find that learning to properly use scikit-learn and tensorflow (or your machine learning algorithm of your choice) is atleast an order of magnitude lower of work than getting the data. For instance, I wanted to build a machine learning algorithm which took simple data from the SEC filed 10-Q and 10-K, which are…
Because there is a real moat with data ownership and pipelines. If you want to do any analysis you quickly find that learning to properly use scikit-learn and tensorflow (or your machine learning algorithm of your choice) is atleast an order of magnitude lower of work than getting the data. For instance, I wanted to build a machine learning algorithm which took simple data from the SEC filed 10-Q and 10-K, which are…
> data ownership and pipelines not ML. By your argument, the money is in data hoarding and brokering, and renting that data (with DRM) to ML outfits, not doing ML. Anyway, cleaning dirty data isn't execeptionally hard, it's just boring work; getting the raw data is the hard part.
Because there is a real moat with data ownership and pipelines. If you want to do any analysis you quickly find that learning to properly use scikit-learn and tensorflow (or your machine learning algorithm of your choice) is atleast an order of magnitude lower of work than getting the data. For instance, I wanted to build a machine learning algorithm which took simple data from the SEC filed 10-Q and 10-K, which are…
So much this!!! In all fairness, it doesn't matter what you pick up, you'll spend north of 80% of your time preparing and pre-processing data, which is almost always catastrophically tedious and boring. Annoyingly that also applies to publicly available datasets - pre-processing is still most of the work. For instance, the tensorflow team invested a lot of time and effort into tf.data for that reason but imo it doesn…
On a side note... data pre-processing is often viewed a side job that needs to get done before the real work can begin. I don't think I've ever been able to prepare a data pipeline without making decisions about the data that will impact the outcome. For example
How do you deal with missing data? Interpolate, ignore, use averages, use a machine learning algorithm to plug the gaps?
How do you decide what data sources to include in the pipeline. What if one data source seems more reliable, but another has far more data, too much to use. Should you amplify it, sample from the other data sets to make the volumes equal, keep things proportional?
What if one data set changes more rapidly than another, how should you update the data set used to populate the ML model?
These are just a few that jump into my head, and while there are techniques to deal with them, ultimately, there isn't a correct answer. And honestly, even these examples make it all seem more glamorous than it is, a lot of this is just figuring out why various encoding and formatting errors are breaking the feed, why column headers mysteriously change, why handwriting on form scans gets properly translated into text some of the time and completely garbled in others.
The funny thing is, I do see people fine tuning ML algorithms (kaggle style, seeing if they can wring a bit more predictiveness out of a model), when in the real world projects, decisions upstream about the data pipeline will have an impact perhaps 5-10 times greater than any tweak to the ML parameters (or even which general algorithm to choose). And yet, it's hard to get people to even pay attention to these decisions, probably - as you said - because people find it catastrophically tedious and boring.
It's like the gold rush except the gold (data) is easy to make, and very very cheap. hmm better sell shovels.
Yeah, people say that, but it's hard to see what selling shovels would amount to. I mean, hardware/cloud is commoditized, ML software is free. Maybe Uber for annotators /s (read, Mechanical Turk). (Of course, there is the ever more popular route of making services for enterprises wanting to outsource basic stuff.)