They didn't use meeting transcriptions or the query features related to that, which in my opinion is the primary value add of Copilot.
I have been using Copilot for over half a year or something, and I haven't found it to be useful in Word, and it is rarely useful in Outlook. I don't use Excel a lot, so I get a lot of value out of the formulae generation stuff when I have to use Excel.
There are several meetings that I need information from, but don't need to attend. Culture has been to take the summary from a meeting and dump it into the chat after, and if it misses anything add it below it. Really nice for action items and things like that.
Idk. I've read through those summaries after meetings and they can often be misleading of miss important context. I still think you need to have a human to read through the meeting summaries to confirm they are accurate.
I really want 365 or gemini to do the obvious thing: "Using this template, summarize this document into a slide deck."
Maybe it won't be good, but it's surprising to me it can't do it barebones. Considering LLMs do an okay at summarizing into markdown, maybe the issue is a lack of proper intermediate DSL to represent slides it can train on?
But autogenerating at least a scaffolding for a deck seems so obvious, it's pretty damning it's so bad 2 years later.
All my documents - Word, PowerPoint, OneNote, Excel - are all on OneDrive. Give me an NLP interface to query information from these sources. When providing the information, note the source where it came from. That's what I want!
When I tried it with a copilot trial at one business, it basically just dumped huge text blobs into big text fields.
They push the "generate image based on prompt" as the main add for slides to make them not just bullets, but no real foundational slide creation is baked in.
Things may have changed since I demo'd it ~6 months ago.
Google Slides is just generated images and transform text as far as I've seen.
We have information and we need to organize it so it can be presented so leadership can make a decision.
We now have so much information that the effort needed to take this step is so obnoxious that need an abstract organization system to exist on top of it.
I wonder if the scaling properties here just serve to highlight the deeper organizational problem that we still haven't solved properly yet.
Whattttttt? Impossible. Surely the AI must have been adding value. A big company wouldn't just look to shove AI into something because it's trendy and fool themselves into thinking that customers would pay them for it, would they?
Also, why are we linking to some forum and not the articles?
Given what people are fantasizing what “AI” can do, it's not surprising MS thought their users must be ecstatic about this. Besides, MS always knows better what their users want than the users themselves.
My boss, the VP of Product, went on a spree about GenAI these past few months. We've been working on regular prediction and optimization things, but when ChatGPT went nuts, everything was GenAI. He was reading generative AI into RFPs and proposals that didn't even mention generative AI.
I tried talking to him that it's still garbage in and garbage out. If the solution was never there to train on, LLMs aren't going to make them out of thin air. There's also the "language" part of LLMs, that he didn't get.
We get pulled into a workshop with Amazon, who also wouldn't give us free resources unless it had a GenAI angle. After all this effort, you know what we made?
A fucking chat bot.
A fucking chat bot, that's also near useless, because our customers are highly driven and only interested in mathematical predictions. Turns out, you have to actually solve the hard math problems first with many examples, before GenAI can even take a terrible attempt at it.
They gave my brother in law’s company a discount on the yearly renewal if they enabled GitHub copilot for all 5000 devs or so in the organization. Our theory was they thought all the devs would become addicted to copilot then demand the company pays for it the following renewal.
Teams used to be one button to attach a file to a chat.
Now I have to hit the '+', at the bottom of the window, wait for the dialog to fully load, then click 'Attach file', at the top of the window, wait for the dialog to load again, then click 'Upload from this device', again at the bottom of the window.
I sincerely HATE my coworkers for asking for attachments now.
No one at Microsoft is paying attention to the product as a whole. It seems like teams of middle managers all vying to get their dumb feature included in the product. It's an incomprehensible mess of poorly thought out misfeatures.
I still haven’t figured out how to export a copy of an excel sheet to csv without that csv becoming the document I have open afterward, and having to close it and reopen the xlsx file.
(note: desktop Excel 2016)
There's an "Export" section in the awkward full-window UI that they replaced the File menu with, but instead of exporting it's just another way around to add extra clicks (File -> Export -> Change File Type -> CSV -> Save As) and get to the same "Save As" dialog that makes the CSV file take over your window as the active file after you "export" it.
You can also get to that via Save As -> Browse and pick CSV in the "save as type" dropdown, which behaves how you expect Save As to work (where that copy should stay active in the window afterward, as opposed to Export where it shouldn't). But this spot is also stupid because now have to go File -> Save As -> Browse because somebody at Microsoft wanted to insert an extra page where they could default your location to OneDrive and make you click Browse to get a normal file browser.
I saw an article yesterday about how Microsoft is moving toward "user-centric design" and removing ads from Skype, but how about doing that in the products anybody gives a shit about?
Dumb as it is, the quickest way to get a CSV export is probably to make a new document and copy/paste the contents of the sheet into it, then save as CSV and close
First the inital, irrational exuberance. Along with out of control spending on anything AI.
Then the lack of financial return on investment from the initial exploratory (try everything) stages. Leading to a sizable pullback in spending across the entire ecosystem.
Then the doom & gloom (talking heads on TV; smarter-than-thou industry pundits); death of garbage start-ups (the great thinning); painful layoffs; stocks decline.
And last, given some time - very substantial, tangible value (productivity gains, new products) is realized across decades. Rinse & repeat.
This says more about Microsoft's execution than AI. Their bing chat is a bloated mess and while I haven't personally tried AI Office 365, I suspect it follows the same theme.
Anecdotally, I've heard demand for OpenAI's enterprise tier is ridiculous from someone working within their GMT function. I guess you can say Microsoft did a great job of hedging their bets.
> This says more about Microsoft's execution than AI.
This says pretty clearly that what Microsoft is peddling isn't even remotely an AI. They used AI purely as marketing term and ignored the obvious market consequences that would arise from that.
Why is this post off the front-page after < 1 hour? or even the second page, or third page? In 1 hour, it garnered ~60pts, and it's a very interesting topic.
I seriously think Big AI (Microsoft, Open AI and other players) somehow influenced this burying.
It's super-damaging to Microsoft, who have tooted O365 "AI" like crazy. Not to mention their investment in OpenAI. Continued criticism like this about their clumsy "AI" is probably going to cost them a few billion.
> I seriously think Big AI (Microsoft, Open AI and other players) somehow influenced this burying.
This feels unnecessarily conspiratorial. Occam's razor would imply that it's just that a lot of legit HN users are pretty emotionally invested in this stuff, and dislike seeing negative stuff about it. You can see this in the comments on these sorts of articles; there's definitely a cohort who have a quasi-religious zeal about this stuff.
they usually give it for free, so they can report to markets that there's a strong adoption, then a year or so down the road, they start charging a little bit.
This, combined with the otherwise non-specificity of the article gives me pause.
Is AI bunk? Sure! Even the latest-and-greatest models can't explain how to programmatically create a DNS record in Azure using the latest C# SDK, how to create a searchable/sortable React data-table with a remote data source, or how to achieve world peace (just to cite 2/3 topics everyone utterly failed at the last month when I tried).
Is this article bunk? Yeah, most likely -- no exec in any pharma company in their right mind would admit failure after only six mere months, and without any lawsuits...
I have been using Copilot for over half a year or something, and I haven't found it to be useful in Word, and it is rarely useful in Outlook. I don't use Excel a lot, so I get a lot of value out of the formulae generation stuff when I have to use Excel.
There are several meetings that I need information from, but don't need to attend. Culture has been to take the summary from a meeting and dump it into the chat after, and if it misses anything add it below it. Really nice for action items and things like that.