Problem is you really don't even need ML for the curve fitting tools. Just ARIMA or Holt Winters is enough and likely faster. Heck, Kubernetes HPA doesn't even implement full PID, just P.

It seems like the main challenge is going to be finding domains where accuracy is important enough to spend inference token money on, but not so important that you'd hire a team to build a custom model. Or maybe that team borrows Timestral weights to fine tune to the domain, but they probably already have a different model customized to the problem?

After that they probably did a security sweep and found the active one.

This is more or less how I already set up GNUCash. Couple dozen scheduled transactions created 90 days into the future, with the "Future Minimum Balance" field enabled on the homescreen.

For me, the time sinks are interrelated:

1. Monthly reconciliation. Every month I sit down and compare the scheduled transactions with what actually happened in statements, for approximately thirty different accounts. Many don't have APIs, certainly none GNUCash would speak. This takes a lot of time and I wish I could just have this automated. Just give me exception handling on the edge cases like a new transaction category, or a forecasted negative checking balance.

2. CC transaction import. This is a structural failing of banking closing down formerly open APIs. Only thing left on my BofA accounts is CSV exports. Which thankfully works with a bit of setup but the writing is on the wall and even that may soon vanish in the name of banks defending their core business.

3. Stock transactions, equity grants & price tracking. This one is even harder since CSV feels structurally not up to the task of representing the "sell to cover taxes" split of RSU grants, and I never really got it working in OFX anyways. Plus I have a pretty weird collection of accounts including a pension IAP that updates once a year (and that update is retroactive to 6 months ago) and a 457b invested in restricted custom funds with no ticker symbols.

"The flaw of averages" strikes again!

But more fundamentally, SS is more of an age pension[1]. Super is the step above that, akin to forcing all americans to fund their 401k / IRA.

[1]: https://en.wikipedia.org/wiki/Social_security_in_Australia#A...

Where do the agents get tools from? Local artifactory?

[1]: https://appleinsider.com/articles/18/01/09/iphone-battery-fi...

While on the one hand, you do need some kind of grounding in human specification for what to build and what good looks like, any particular defect humans can find should be findable via software.

It started off promising, with some hints about monte carlo methods implemented in spreadsheets, but kind of detours into statistics basics, layman summaries of the author's prior scholarly work, and various business consulting anecdotes. I did not, for example, expect to read a chapter on the FASB-II's flaws regarding options pricing. And a few are downright concerning a decade later. I would not, for example, brag about advising Wells Fargo executives on their employee incentive programs after the cross-selling scandal came to light.

The final chapters, which I have not yet gotten to, supposedly cover the solution to the flaw of averages. Given the book is 15 years old, I'm expecting it to be pretty dated implementation wise, but I might be able to apply it to Prometheus histograms or t-digests. And I've learned a few things, like Jensen's inequality, and found a few sheets demoing a sampling approach.

So why do it? Because datacenter regulations on earth are becoming a referendum by proxy on AI. "Texas has no jurisdiction in low earth orbit!" is the general idea. Critics point out that this is a transparent move to moot the argument over AI, and that it doesn't even work in principle, due to above mentioned physics.