Are you kidding? Microsoft has always been shit at math. According to Microsoft Excel, 2 + 2 = 12:04 AM Jan 1, 1900.
this post was submitted on 21 Aug 2025
697 points (98.1% liked)
Technology
74324 readers
3767 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 2 years ago
MODERATORS
Integers are days in Excel, no? So I think 2+2= 12:00 AM Jan 5, 1900.
Intel already did that in the 90's with the FDIV bug.
Wrong, they already had that with Excel. There were a bunch of functions that delivered wrong returns for years, and none of the users (mostly economists) had noticed.
What, you don't always work with 16 digit numbers that are automatically truncated? What could go wrong? We don't use 16 digit numbers for anything, really./
It's hard to believe that's still a thing but it is!
Give Microsoft some credit! Excel has been able to come up with wrong answers for decades. For example, reporting 1900 as a leap year.
That was partly a result of seeking explicit compatibility with Lotus, IIRC.
seeking explicit compatibility with Lotus
I need a shower.
This is such a misguided article, sorry.
Obviously you’d be an idiot to use AI to number crunch.
But AI can be extremely useful for sentence analytics. For example, if you’re trying to classify user feedback as positive or negative and then derive categories from the masses of text and squash the text into those categories.
Google Sheets already does tonnes of this and we’re not writing articles about it.
Yeah, it's like complaining that a hammer isn't good at turning a screw. There's a whole trend of Chess content creators featuring games against ChatGPT where it forgets the position or plays illegal moves, and it just doesn't mean anything. ChatGPT was never designed or intended to be able to evaluate a chess position, and incidentally, we do have computer programs that do exactly that and have been better than any human player since the 1990s. So what is even the point?
And what you could do is to enable an LLM to use these tools and reason about their outcome. Complaining that an LLM isn’t good at adding numbers is like complaining that humans aren’t as fast as calculators when multiplying large numbers.
There's an old story about the lead developer at Texas Instruments saying "I want a computer that fits in my pocket". And then his staff dutifully measured the pocket to spec before proceeding to perform a feat of miniaturization that would revolutionize the modern world.
I'm trying to imagine one of the techies, from way out in the back, saying "Does it have to get the right answer?" Then getting fired, walking off the job, and walking into Microsoft with 10x the salary the next day.
Why would anyone use an LLM as calculator?
That just doesn't make sense.
It is like using a calculator as typewriter because it can spell 80085.
So what you are saying is, my car is a typewriter?
Did you just take a picture of your car's boobs at 60k/h? High speed boobs shots hahahaha
Maybe :P
It's a legal requirement when your car hits 80085 that you must take a photo, it supersedes all other laws.
load more comments
(4 replies)
To waste electric energy. All those power plants produce immense amounts of energy that needs to be consumed. If we didn't have LLMs, the pollution of those plants would be for nothing. At least now, there is an attempt to put it in good use.
load more comments
(3 replies)
There are things that could be done to improve Excel. For instance, fully integrate python and allow it to be used to create custom functions. Then, maybe one day, VBA can ride off into the sunset where it belongs.
Adding Copilot to Excel is not an improvement because Copilot and all other LLM based platforms frequently barfs out totally incorrect information about how to do something in Excel.
"You do that using formula."
No, I can't, you worthless pile of shit because THAT FORMULA DOESNT EXIST.
Integrated python scripts in excel sounds like a malware developers dream.
I mean... Yeah, but the same can be said for VB?
Especially since VBA can make calls to the Windows API directly and through that avenue do all kinds of funky things to your system.
load more comments
(1 replies)
load more comments
(20 replies)
load more comments
(30 replies)
"Microsoft Excel is testing a new AI-powered function that can automatically fill cells in your spreadsheets."
Every year, Microsoft gives me more reasons to permanently leave their products.
Unfortunately, due to compatibility with financial and other Windows-only software I still need to run Windows, but I am down to two rigs and it might go down to one in the new year.
A virtual machine running windows, to host just those apps, might be a good step away at this point.
load more comments
(2 replies)
Our very own economic Butlerian jihad.
Such a complicated way just to add more RAND() to formulas.
OK, I'm not really mad at this. I already used Copilot to design a table for me in Excel and it worked really well. It did everything for me, and I just had to copy-paste the formulas into their appropriate spots. If it's built-in, possibly will work better.
Not everybody needs to be an Excel expert, after all. Having that functionality might be actually beneficial.
How do you know those formulas are correct?
By verifying that they're correct...? 🤔
I think the concern is that you can come up with a number of formulas that will get correct answers for some combinations of values and not others.
If you do not understand the logic of the formula, and what each function does, how do you verify they are correct and will always give you the results you think they will? Double check every result in its entirety?
That's my thinking
If you know what you're doing, it's significantly easier to do it yourself
You at least have some reassurance it's correct (or at least thought through)
Verification is important, but I think you're omitting from your imagination a real and large category of people who have a basic familiarity with spreadsheets and computers, so are able to understand a potential solution and see whether it makes sense, but who do not have the ability to quickly come up with it themselves.
In language it's the difference between receptive and productive vocabulary: there are words which you understand but which you would never say or write because they're part of your receptive, but not productive knowledge.
There are times when this will go wrong, because the LLM will can produce something plausible but incorrect and such a person will fail to spot it. And of course if you blindly trust it with something you're not actually capable of (or willing to) check then you will also get bad results.
yes okay dad
Any time, nutsack.
I'm a dad and I approve this message.
view more: next ›