4 comments

[ 2.6 ms ] story [ 14.6 ms ] thread
LLMs are on the verge of replacing data scientists and investment bankers. But can they perform simple accounting tasks for a real business?

We built AccountingBench, a test where LLMs must "close the books" for a real SaaS business using 1 year of Stripe/Ramp/Rippling/Mercury data.

Claude 4 and Grok 4 start strong - within 1% of human CPA baselines in month 1.

But as time progresses, all models inevitably accumulate compounding errors and exhibit erratic behavior, causing significant deviations.

That said, the early accuracy here is promising. With targeted post-training, models may be able to replace humans for this kind of work.

Given their inclination to fabricate user-pleasing answers...could I let an LLM do my tax returns?
I was just thinking of that earlier today, really cool!
LLMs are really not good at following specific processes like math. They operate off vibes.

Ask Claude to multiply two ten-digit numbers. It gets the first one or two digits correct, and then makes up the rest.

ChatGPT used to have the same problem, but now it writes a program to perform the math for it.