FR version is available. Content is displayed in original English for accuracy.
Advertisement
Advertisement
⚡ Community Insights
Discussion Sentiment
58% Positive
Analyzed from 5276 words in the discussion.
Trending Topics
#excel#copilot#add#claude#microsoft#chatgpt#model#sheets#more#don
Discussion (166 Comments)Read Original on HackerNews
I recently tried Claude Cowork for PowerPoint and I was stunned by the content as well as design quality of the deck it produced. That's a threat for Microsoft because now you don't need the editing tools of PowerPoint, AI replaces it, so all you need is the presentation mode of PowerPoint.
Copilot for Excel is useless. Ask it what is in cell A1 and it can't answer. I am looking forward to trying ChatGPT for Excel.
Claude for Excel (I work in finance) was one of the absolutely critical reasons we added Anthropic enterprise licenses. But they've turned out to be quite expensive ($100/day for heavy users). We'll see what OpenAI's quotas are.
You get models that are formatted and structured and which balance - but there are errors introduced which an analyst / human wouldn’t make.
Stuff like hard coded values, or incorrect cell logic which guarantees the model balances.
I don’t get good results when I just have Claude build things on its own - but for these types of specific productivity tasks I can save a couple of hours here and there.
Does this remove (or at least increase) the upload limit?
* A large application whose outputs are independent of the all (people still print slides; when presenting nobody knows or cares what app was used) * Complicated and requires users to learn lots of skills unrelated to the work they’re doing (compare to Excel, where the model and calculations require and reflect domain knowledge about the data) * Practically zero value add in document / info management (compare to word where large documents benefit from structure and organization)
We’re pretty close to presentations just being image files without layers and objects and smartart and all that.
AI will come for all productivity tools, but PowerPoint will be the canary that gets snuffed first, and soon.
HN discussion: https://news.ycombinator.com/item?id=47603231
That’s without talking about the poor UI and security story of COM add-ins and the inability to run on Excel for iOS.
I was hyped when I heard about Copilot. "I can tell it to make pivot tables now!" When I tried to use it I was shocked how underbaked it was. Below even my worst expectations. This really was someone shoving ChatGPT into Excel with almost zero additional effort. Copilot can't DO anything useful.
I am surprised that Microsoft's own copilot product is so far behind though.
Maybe(?) from a product catalogue perspective... But from a strategic perspective less so because they own ~27% of OpenAI.[0]
[0] - https://openai.com/index/next-chapter-of-microsoft-openai-pa...
The MCP ecosystem is what makes this interesting. Claude isn't just a chat panel bolted onto existing software, it's building integrations that actually manipulate the files. Microsoft had the distribution advantage but they're losing on capability.
Regardless, I just tried to log in with my work MS account, and I can't do so.
Microsoft fumbled so badly here.
Actually, someone here posted a Claude Code skill recently that generates a presentation as a self-contained HTML5 file, so all you need is a browser.
PowerPoint, as a whole, is doomed.
My employer blocks office plugins, so I can't try Claude for PowerPoint, but sometimes I get Claude to generate Python scripts, which produce PowerPoint slides via python-pptx. This also benefits from being able to easily read and generate figures from raw data.
I don't really like the way Claude tends to format slides (too much marketing speak and flowcharts), but it has good ideas often enough that it's still worth it to me. So I treat this as a starting point and replace the bad parts.
we've hit this point where its cool to have claude reinvent every wheel just because it can.
For added benefit, full screen?
Until you need presenter notes or other niceties, this covers a large space of usage.
Nowadays I just make single-purpose websites with Claude Code because Google Sheets has such poor AI integration and is outrageously tedious to edit.
They had all the parts and I have a subscription and it still does terrible things like prompt me to use pandas after exporting as a CSV. It will mention some cell and then can’t read it. It can’t edit tables so they just get overwritten with other tables it generates.
It reminds me of something a friend told me: he heard that Google employees do dogfood their products; some even multiple times every year. There’s no way anyone internal uses Sheets even that often.
Given how well the API works, that we are discussing Googlers, my guess is that's how they dog food their services. Programmers don't get hired by Google for mouse skills.
The GUI is for spot checking results, final presentation.
If you're sitting there point-n-clicking everything into place perhaps consider you are doing it wrong.
I have no idea if they do or not, but it's a plausible explanation...
Do you mean restricted workflow? Googles APIs are pretty much 1:1 to the GUI
And using Python makes it trivial to copy-paste out of files and other APIs with one run of Python
Versus all the fiddling in browser tabs with a mouse, it actually affords an incredibly wide set of options to quickly collate and format data
The libraries themselves are OK, but MS uses them stupidly. If you want to fill out some form in DOCX or XSLX format you will get broken formatting. And this is from Office company.
Claude one-shot something with a Python script that was pretty okay.
We came up with what I still consider a pretty cool batch-rpc mechanism under the hood so that you wouldn't have to cross the process boundary on every OM calls (which is especially costly on Excel Web). I remember fighting so hard to have it be called `context.sync()` instead of `context.executeAsync()`...
That being said, done poorly it can be slow as the round-trip time on web can be on the order of seconds (at least back then).
If you were working on the platform itself, then I would be interested in hearing your more detailed thoughts on the matters you mentioned (especially since I am developing an open source Excel Add-In Webcellar (https://github.com/Acmeon/Webcellar)).
What do you mean with a "OM" call? And why are they especially costly on Excel web (currently my add-in is only developed for desktop Excel, but I might consider adding support for Excel web in the future)?
In any case, `context.sync()` is much better than `context.executeAsync()`.
The reason those calls are expensive on Excel Web is that you're running your add-in in the browser, so every `.sync()` call has to go all the way to the server and back in order to see any changes. If you're doing those calls in a loop, you're looking at 500ms to 2-3s latency for every call (that was back then, it might be better now). On the desktop app it's not as bad since the add-in and the Excel process are on the same machine so what you're paying is mostly serialization costs.
Happy to answer more questions, though I left MSFT in 2017 so some things might have changed since.
Regardless, I have always preferred Excel desktop over Excel web (and other web based spreadsheet alternatives). This information makes me somewhat less interested in Excel web. Nonetheless, I find Excel Add-Ins useful, primarily because they bring the capabilities of JavaScript to Excel.
Honestly, I struggle to think about what has actually changed between Office 2013 and Office 2024 (and their Office 365 equivalents); I know the LAMBDA function was a big deal, but they made the UI objectively worse by wasting screen-space with ever-increasingly phatter non-touch UI elements; and the Python announcement was huge... before deflating like a popped party balloon when we learned how horribly compromised it was.
...but other than that, Excel remains exactly as frustrating to use for even simple tasks - like parsing a date string - today just as it was 15 years ago[1].
[1]: https://stackoverflow.com/questions/4896116/parsing-an-iso86...
What is the data model that you use for the spreadsheet itself? I found I could create a chat completion persona that believed it is one of the developers of a popular open source spreadsheet, and I put this "agent" directly inside the open source spreadsheet. I did this before tool calling was available at all, so I made my own system for that, and the "tools" are the API of that open source spreadsheet. My agent(s) that operate like this can do anything the spreadsheet can do, including operate the spreadsheet engine from the inside.
I made a CLI (+skill) so agents could edit files with verbs like `insert A1:A3 '[1,2,3]'`, but did some evals and found it underperformed Anthropic's approach (just write Python).
It's here: https://github.com/tmustier/pi-for-excel
This time even for pro.
Tune in for the next episode: Word
I asked GPT-5.4 High to draw up an architecture diagram in SVG and left it running. It took over an hour to generate something and had some spacing wrong, things overlapping, etc. I thought it was stuck, but it actually came back with the output.
Then I asked it to make it with HTML and CSS instead, and it made a better output in five seconds (no arrows/lines though).
SVG looks similar to the XML format of spreadsheets. I wonder if LLMs struggle with that?
I just spent a few hours trying to get GPT5.4 to write strict, git compatible patches and concluded this is a huge waste of time. It's a lot easier and more stable to do simple find/replace or overwrite the whole file each time. Same story in places like Unity or Blender. The ability to coordinate things in 3d is really bad still. You can get clean output using parametric scenes, but that's about it.
This seems like a security nightmare, which is especially relevant because sensitive data is often stored in Excel files.
It seems one of the biggest barriers to people's adoption is concern over data leaving their ecosystem and then not being protected or being retained in some way.
Is this is an SLA that a small or medium sized company could get?
However, it may be important to note that these security considerations are relevant for most Office Add-Ins (and not just the ChatGPT add-in).
This is counter to the old (security nightmare) COM model where processing could be local.
> (…) you can verify each step and revert edits if needed.
I wish there were different workflows.
It feels like current most popular way of working with GenAI requires the operator to perform significant QA. The net time savings are usually positive. But it still feels inefficient, risky and frustrating, especially with more complex and/or niche problem areas.
Are there GenAI products that focus more on skill enhancement than replacement? Or any other workflows that improve reliability?
Just this past week I used it to generate a simple model of a few different scenarios related to an investment property I own.
The first problem I ran into is that it was unable to output a downloadable XLS file. Not a huge deal - it suggested generating CSV tables I could copy/paste into a spreadsheet. The outputs it gave me included commas in a handful of numbers over 1,000 (but not all of them!) which of course shifted cells around when brought into Google Sheets. We pivoted our approach to TSV and solved this problem. Big deal? No. Seemingly basic oversight? Absolutely.
This is where the real fun began. Once I started to scrutinize and understand the model it built, I found incorrect references buried all over the place, some of which would have been extremely hard to spot. Here's my actual exchange with ChatGPT:
- - - - - - - - - -
> Can you check the reference in cell F3? It looks like it's calling back to the wrong cell on the inputs tab. Are there similarly incorrect references elsewhere?
> Yes, F3 is incorrect, and there are multiple other incorrect references elsewhere: (It listed about 30 bulleted incorrect references)
Bottom line - - Many formulas point to the wrong Inputs row because of the blank lines - The Sell + Condo section also has a structural design problem, not just bad references.
The cleanest fix is for me to regenerate the entire AnnualModel TSV with: - all references corrected - all 15 years included - the condo scenario modeled properly with a separate housing asset column
- - - - - - - - - -
This was me asking about the exact output I had just received (not something I had made any changes to or reworked.)
There are plenty of domains where I have enough faith and error tolerance to use ChatGPT all day, but this just sends a chill down my spine. How many users are really going to proof every single formula? And if I need to scrutinize to that level of detail, what's the point in the first place?
You're not wrong, but you'd think that given their 27% stake in OpenAI they'd put more weight behind ChatGPT integration.
Damn that OAI valuation is like a sore boil that is about to explode.
Also once again, a lack of imagination from OAI. Damn vision really is super scarce huh.
Obviously doesn't apply to everything, and there are some features that are very hard to replicate. But still.
meanwhile not that ant is genius, except the timing of dow drama right before Iran war.
Instead of answering with 6, it came up with 15. The comment was "If AI is doing this, a global financial crash is inevitable."
Might not be real but it is something to keep an eye on. Hopefully, they are a bit more cautious on how this is implemented.
Show HN: I've built a C# IDE, Runtime, and AppStore inside Excel
670 points | 179 comments
One of the main use cases was to analyze Excel data with SQL. I'm the kind of nerd that loves stuff like that, but stuff like that seems completely obsolete now.
[0] https://news.ycombinator.com/item?id=34516366
Building an agent that can securely access systems of records, external data sources, and other files in your workspace—with context for the work you do outside of Excel—is where the revolution is at.
I would never use Copilot for anything useful, but I do use OpenAI products.
It doesn't matter when you use something else wholesale under the covers, if you botch the token spent...
Microsoft, being Microsoft, will find a way to win no matter who that vendor ends up being.
So, yes but no. Not that I care, but the answer to the above question is a no, and should start with No.
There are now just even more errors than there already were.
Now there's hope though: I take it at some point, just like we have AI that can already find (and fix and sometimes even properly fix) errors in code, we may end up with AI tools able to find all the broken assumptions and errors / wrong formulas the spreadsheets that make the corporate world are full of. But atm that's not where we are.
One such corporate-world company producing a gigantic turd would the "biggest" (but it's really not that big) european software company, SAP... They're going full on "business AI" as they see (rightly so?) AI as a terminal death threat to their revenue model. Market cap went from $360 bn to $200 bn: don't know if it's related to their "genius" AI-move.
And so now we have countless corporate drones who were already incapable of doing any kind of financial/accounting/math computation in a rigorous way who are now double-speeding on the errors, but this time AI-augmented.
It's the "let's add an AI chatbot to our site" (which so many companies are adding to their websites right now), but corporate version: "let's add AI to our corporate tools".
Just to be clear: I think this cannot fail. Failure and bogus numbers are the norm in spreadsheets, not the exception. More failure, more bogus computations, actually won't change a thing.
They (OAI+Anthropic) very much do not get exactly what these people are doing in the job (accounting+corporate finance+valuation+asset management) and what the actual production process is. These tools are irrelevant, disrupt flow and if anything just add noise to what one is doing.
They don't do the job, reliably or well. No amount of wishful thinking or extra tokens will change that.
Remember when Steve said 'The computers for the rest of us'?
I suppose it isn't a surprise. Are researchers/generally geeky people meant to be able to relate to the average person's day-to-day beyond their sphere? Lmao.
You can't produce stuff for people you don't understand. Understand being a very key term.
I know there are employees of those firms here that would love to know. But nah lmao.