Sure downside would be not learning from people using your model for coding, if we're on the cusp of huge leaps in self-improvement. But there is a reasonable case for avoiding desperate scramble, especially if other parts of the business can also create value with the compute.
Version 3.1 has plenty of room for improvement, yet they don't seem to be giving the attention it deserves or at least communicating accordingly.
It may be that they wish to slow their cadence of releases, or develop their models to focus more in a different direction, etc. No matter what the actual reasoning, they have chosen to not compete in the same race, and I cannot say I fault them.
They never gave an official answer as to why, so I'll let you draw your own conclusions.
They did not decide it wasn't worth spending the money to train.
They absolutely spent the money.
Also, look at Flash 3.5 to 3.7. Flash 3.7 is a genuinely decent Sonnet 5 class model. Flash 3.7 is quite efficient too. Also, whatever was spent training 3.5 pro is probably not wasted. However, as a strategy, when I see models like Kimi K3, Fable, Sol. If you discard "because the model sucked" what other alternatives or potential options might exist?
I thought of a quite a few and they are far more compelling and interesting to me.
(Also Gemini models tend to be pretty decent at more than just programming. Enterprise AI use is more than just software eng / programming)
If you'd told me at the end of Cloud Next 2025 that by now Google still wouldn't have a competitive offering to agentic coding offerings from Anthropic (Claude Code + Fable) or OpenAI (Codex + Sol), I wouldn't have believed you.
In our non-coding use cases where we're embedding models in our product, we're also not reaching for GCP stuff. Because Anthropic has the mindshare of our engineers and product folks, since it's what they use every day.
They mentioned that they have already started pretraining Gemini 4, which will be the full ground up rip-your-face-off-expensive training that is often discussed.