3 ms·

Open source also serves as a "defense strategy" for larger companies whose core business does not directly / indirectly depend on commercializing its open sourc

Open source also serves as a "defense strategy" for larger companies whose core business does not directly / indirectly depend on commercializing its open source project(s). They in turn serve more like hedges. For example, Meta with PyTorch in context of the machine learning tooling space. PyTorch is probably strategically important for them as they do not have a cloud business. The other large competing deep learning framework is Tensorflow, designed and open sourced by Google but also owns a cloud business (GCP). They also have in-house AI accelerators like TPUs. Microsoft didn't really have a deep learning framework, but they have ONNX, and is now using ONNXRuntime as an on-ramp to their cloud business (Azure). If Meta wants to continue doing ML, and all the other cloud players jacked up their prices, that's gonna hurt Meta's operating costs. So it's still "cheap" to have hedged against such a scenario. All these on top of upsides like developer marketing etc.

mousetree 3y ago
> If Meta wants to continue doing ML, and all the other cloud players jacked up their prices, that's gonna hurt Meta's operating costs How so? Meta is not running AWS/GCP/Azure.
- tzhenghao 3y ago
  They’re not, but there’s non-zero probability most R&D below a deep learning framework gets fully vertically integrated (or even aggregated in the short term) to just the big cloud players.
  - indymike 3y ago
    Give this time, and it will change.
bootsmann 3y ago
Open sourcing Pytorch was also a human resources decision. Big tech learned this lesson somewhere around 2014, that whatever problems they solve, other companies will have 5 years down the line and if they don’t open source their solutions someone else will release the industry defining standard and meta & co will have trouble hiring people for their handrolled solution 15 years down the line. IIrc some spotify engineer said in an interview once that this was a big reason behind them open-sourcing backstage.
satvikpendem 3y ago
This strategy is called Commoditizing Your Complement [0]. [0] https://www.joelonsoftware.com/2002/06/12/strategy-letter-v/ https://www.joelonsoftware.com/2002/06/12/strategy-letter-v/

mousetree 3y ago

> If Meta wants to continue doing ML, and all the other cloud players jacked up their prices, that's gonna hurt Meta's operating costs How so? Meta is not running AWS/GCP/Azure.

tzhenghao 3y ago
They’re not, but there’s non-zero probability most R&D below a deep learning framework gets fully vertically integrated (or even aggregated in the short term) to just the big cloud players.
- indymike 3y ago
  Give this time, and it will change.

bootsmann 3y ago

Open sourcing Pytorch was also a human resources decision. Big tech learned this lesson somewhere around 2014, that whatever problems they solve, other companies will have 5 years down the line and if they don’t open source their solutions someone else will release the industry defining standard and meta & co will have trouble hiring people for their handrolled solution 15 years down the line. IIrc some spotify engineer said in an interview once that this was a big reason behind them open-sourcing backstage.

satvikpendem 3y ago

This strategy is called Commoditizing Your Complement [0]. [0] https://www.joelonsoftware.com/2002/06/12/strategy-letter-v/ https://www.joelonsoftware.com/2002/06/12/strategy-letter-v/