Came here to say that. I believe that AI assistance will make a big difference to many professions going forward—certainly any profession that has to deal with large volumes of information, whether it's education, knowledge, or red tape. But I wouldn't conflate that with copilot. I hope that open-source AIs eventually occupy that role.
Imagine the vendor lock in dark patterns to counter competitive AI assistant utility... Even shittier docs, etc.
Just had to use ChatGPT to get help about ASP.Net APIs as the docs are way worse than used to be 10 years ago, and are full of sales plugs for products instead of explaining the basics/architecture/workflow... MS docs used to be quite okay in my experience. (ChatGPT managed to cut through the bullshit and showed me the APIs I needed but were deliberately not mentioned/detailed in the official docs)
I found GPT-3 significantly better than 3.5, at least initially, for what it's worth. The main benefit of 3.5 was that it was ten times faster/cheaper, not that it was better.
*An encrypted copy of Microsoft Bob was included on Windows XP discs to discourage piracy by consuming more space on the CD and consuming more modem-limited bandwidth.
I would presume an Excel AI is trained or fine-tuned on Excel, yes. Anyway, you're not asking these questions in good faith. You've already made up your mind that AI is bad: https://news.ycombinator.com/item?id=37714113
Is there no other word than Copilot? I swear I've seen like 6 AI things named Copilot so far haha. Two are from Microsoft!
I'm open to the idea of voice assistants being useful with a sprinkling of AI, but not if I have to wait 10s+ for a response. They'll at least need to come up with a "waiting on AI" mouse cursor or something to show it's doing something. VS Code Copilot is kinda hit or miss whether it's thinking on the prompt (comments) you gave it or if it has nothing to add.
> Think of CoPilot as having your very own ChatGPT seamlessly integrated into all your Microsoft applications – Windows, Edge, Excel, and PowerPoint.
> In fact, it’s essentially Chat-GPT, powered by OpenAI, but with a unique twist – it’s context-aware and fully capable of interacting with your data and the software running on your PC.
Asking Excel to "add a column averaging the last 3" sure seems a lot easier and faster than having to read the manual to figure out how to do such a thing. This could drastically simplify user interfaces.
It could also confuse users once more complexity arises because people often ask for contradictory things when the constraints of the spreadsheet are not guiding them.
I can do an average function in Excel. But I can not remember the word of the function in all three languages that Excel have decided that I should be proficient in Excel in. The translation of functions is the single worst design decision in that product.
For a contrast: I use Copilot in vscode a lot. There it's pretty good at guessing what I am trying to do, and offering an auto-complete for it.
So mostly it feels like it's just helping me do the thing I wanted to do anyway, but faster.
But if eg a user knows how averaging of numbers works, but just doesn't know how to express it in Excel, even that example would be useful. I agree that if a user doesn't know how averaging works on a conceptual level, copilot won't save their butt here. (At least not yet.)
Welcome to the wonderful world of function names being translated to the installed language.
My first language is Norwegian, and except for a few that I recall, I constantly have to refer to a website like this[1] to figure out what the hell they decided to call a function in Norwegian.
Copilot will collapse for any query more complex than 3 clicks. And any intermediate excel user will be able to perform these clicks quicker than to formulate their prompt.
The second one, probably. But ChatGPT-4 can handle surprisingly complex tasks well for SQL, I don’t doubt it would do well across the office suite.
Obviously it doesn't have the context window to handle things I really want it to do in Python (“refactor this spaghetti mess I got myself into, separate the giant classes into multiple separate modules, fix any O(n^2) parts, add unit tests and do some fuzz testing on function arguments”) but it will be incredible once it does in 10 years or whenever GPU’s have enough VRAM or we figure out how to stuff large context windows into less VRAM.
"Oh.. but skip cell A.. 6? or whatever, the one with the .. uhh... "
'Error processing your request. I emailed your supervisor with my error condition.'
"No, no! Stop. Just do whatever you did first, but now you do like the other thing too, with the skipping of the cell with the.. uh.. what's it called?! ampersand?"
'I will average the last 3 columns while adding random numbers to the rest of your document, sure. By the way, this is the information you requested. I have also forwarded it to the company CEO: 'The ampersand, also known as the and sign, is the logogram &, representing the conjunction "and". It originated as a ligature of the letters et—Latin for "and"'. SOURCE: wikipedia.'
Proceeds to mess around with LLM for 3 years instead of learning wat AVERAGE() is and how it works
I have learned this from experience. If you think this is bad, real people in real situations are actually worse. If you can get them to speak quasi-coherent sentences that’s a major win in itself. Sounds sad, but the reality these gadgets must operate under.
It’s not that people are stupid, but they operate under a surprising amount of varying conditions ranging from fully inebriated to severely demotivated and/or burned out.
It's not as simple as just using AVERAGE actually. I've had the exact problem with Google Sheets. I wanted to create a trailing average of last 7 of the numerical values in another column. I had to Google quite a bit to figure out how to handle it so the 1st to 6th row would work out well. Right now I described the issue to ChatGPT, and it was very quickly able to suggest me the following for Excel:
If you know everything by heart, it may seem simple, but it's not just using one function, it's using multiple and knowing the exact parameters, intricacies.
With Google Sheets, it would've faster for me to just use the JavaScript with Apps Script, to create this, rather than trying to figure out the syntax of this.
So it uses 4 different functions there,
1. IF
2. ROW
3. AVERAGE
4. OFFSET
And in such a way that I would just not come to think of it on my own out of the blue. I would try several different approaches before I were to try this one. Although it's really easy to just write a script for that.
And maybe you also want to be able to dynamically configure the trailing average size, which this formula allows for very easily. Compared to when you Google, some of the articles suggest just using AVERAGE(A2:A8), but it's missing those other frequent requirements.
And when you look at some Stack Overflow answers the formulas are even crazier:
It's odd to me how office people are totally fine wrangling with this mess, but are terrified of programming. This looks way more confusing, magic and brittle to me.
I think the dream goes beyond just that sort of simple "do something in this spreadsheet because I can't be bothered to think".
I think the end game is you open excel and say "Work out what our budget is for the FooSnaggit event based on what Tim said in his email yesterday, and what we spent last year." (for example) and it will go off and read your email and read your other sheets and generate the right thing based on that prompt and that context from your other documents.
I think that will be incredibly powerful ... if it can be trusted. Trouble is now I can't trust these things to get these sort of things correct, so you end up having to double check everything and make corrections etc so you are not really saving that much time.or effort.
These tools can be great for discovery and speed up one's learning curve but there is no replacement for knowing your tools if you're going to master your craft.
In a way, after you devote many years of your life mastering your toolset, you become a very efficient intelligent machine.
But you do also so make mistakes and you also are told to do things by somebody else who guess what didn't spend their life mastering the skills you mastered (otherwise they wouldn't be paying for your services).
From their point of view, you're also some kind of opaque machine that can be trusted to some extent but must be double-checked too. You also can rat them out if they ask you silly things and you also expect to be treated decently as a human being and that's not something many other humans know how to do well.
Unsurprisingly many people cherish the possibility of having to talk to a non-human entity and let them do some tasks they wouldn't know how to even approach. Especially if this entity can be pestered with questions without ever being annoyed.
So far all my interactions with AI in production have involved me trying to figure out the right prompt to be able to talk to a real person who could then actually solve my problem rather than offer me useless text and no concrete action towards solving my issue.
You can't trust an intern with that command either, yet that's exactly the type of task an intern would be assigned today in many large groups.
Of course you'd want to verify the report, but verifying it will probably be simpler & faster than building it, for many tasks. And AI can help you verify as well.
Well for large enterprises that already have their data in the Microsoft ecosystem (Exchange email, Sharepoint etc), Copilot has access to that pool and it's supposed to be isolated within your own environment.
Will output quality of an average excel user go down, if the job duties will evolve into verifying LLM output and not performing low-level tasks over and over?
>>I think that will be incredibly powerful ... if it can be trusted. Trouble is now I can't trust these things to get these sort of things correct, so you end up having to double check everything and make corrections etc so you are not really saving that much time.or effort.
This is the thing for me. I hear about people building apps and stuff but I can't get a working correct script from ChatGPT. Maybe it's me, maybe it's improved, IDK, but the last few times I tried to grab a script for something I was working on, that script had multiple errors.
Yes and no. It might be faster the first time, maybe not so much the consecutive times. Especially when the AI assistant is not doing what you intended.
And in places that have an undo-stack stakes will be quite different from, say, self-driving death machines on public roads.
Someone was already mentioning clippy, presumably as a joke (a well deserved one, I'm old enough to remember). But would an LLM-powered clippy be a joke? Could be the perfect capability/requirements fit.
While I understand that people here think mainly of Github Copilot. The article is actually about Copilot which is them integrating LLM in windows. It is currently available on windows preview versions.
A lot of dumbasses are using the "Copilot" name long after GitHub introduced theirs. Not sure if they're just lazy and unoriginal, or if it's specially intended to cause confusion and get a jump start on publicity.
I don't know about Copilot itself but I agree in general large language models will also bring a huge change like computers, internet or smartphones did. Large language models are still in their infancy as they get better and more people learn how to use them. They will bring a huge change to how computers/smartphones are used by the majority of the people not just the tech inclined. People that mostly use their computers/smartphones just tiktok Facebook and YouTube etc will use it for a lot more as the interface for computing for will change dramatically.
I think biggest change will be how people make stuff so it will impact more in work than in personal space. Think something like if smartphone is car, LLM is more like truck.
It'll probably be as bad as Google, Stack Overflow, or every other thing that people's have complained about making us dumb. Even books and the printing press went through this phase where everyone was worried that the next generation of scholars wouldn't be able to memorize books.
The problem I see is that no one controls the language model in a transparent way. It's not an open database like Stack Overflow, Reddit or other websites that you use when searching for a programming solution, for example.
It has no way of telling what's wrong and what's right. We've seen how AI and bots behave when their training is completely automated. Did everyone forget the bot Microsoft made that became racist after some interactions on Twitter?
I disagree with you as today the sum of all human knowledge is too vast for a group of people to know let alone 1 person. The models as they get better will bring access and result in expanding of human knowledge rather than hampering it. The current models are not very close but I do feel we are moving in that direction
That doesn't mean they're not in their infancy. I take infancy to mean they have not achieved their potential // can be vastly improved or made powerful, as opposed to taking infancy to literally mean 'young' (as in, a period of time).
its funny how most people who complain about copilot dont really know how to leverage it. I NEVER use comments. I should record a video on some tips and best practices
The curious thing is that what looks simple for some might be less obvious for others. I used the old codex models a lot, so I'm aware of a few ways to optimize it's usage. People who learn it in a less optimal way(use comments to say what you want) really start off with the wrong foot
I'm only willing to trust such tools with trivial problems, e.g. I'll ask ChatGPT about some generic snippet I'm sure it's seen in the training data and that would take me 5 mins to type out.
If I have to do prompt engineering to get the model to spit out acceptable output for trivial shit, that's not a model I can trust even for said trivial shit.
"But what I want to focus on is the power of a visionary leader to set out on bold ambitions and bring ideas to life."
Apart from single-product companies I don't think this is really a thing.
CEO's set the company culture, and maybe strategic direction (but it often doesn't need any kind of visionary genious to see which way the world is turning), and of course jump in on any 'big' deals or negotiations the company is dealing with, but are otherwise more curators than creators. IMHO.
I really dislike copilot. I hate how slow it is. The way I write programs is I think about them first. Sketch out on paper if I need to and then start working on it. By the time I am working and in a flow state waiting for copilot to suggest stuff is just annoying. I am baffled when I see respectable people like CTO of Azure praise working with copilot. I feel like there is something I am missing in my workflow
"From a Computer on Every Desktop to Empowering Every Person and Every Organization to Achieve More"
OK, "a computer on every desk" is something concrete. Measurable. Achievable (as in: it is both possible to achieve it, and you can tell whether you have achieved it or not).
What is "Empowering Every Person and Every Organization to Achieve More"? Apart from meaningless corporate gobbledygook?
126 comments
[ 2.5 ms ] story [ 281 ms ] thread(Seriously, what's with the writing in the article? ChatGPT can do better.)
Just had to use ChatGPT to get help about ASP.Net APIs as the docs are way worse than used to be 10 years ago, and are full of sales plugs for products instead of explaining the basics/architecture/workflow... MS docs used to be quite okay in my experience. (ChatGPT managed to cut through the bullshit and showed me the APIs I needed but were deliberately not mentioned/detailed in the official docs)
With Win 11 I just went back to Linux.
We all know current state of AI. It's not ready to control a PC yet.
That would be the first application of AI that I would use daily.
*An encrypted copy of Microsoft Bob was included on Windows XP discs to discourage piracy by consuming more space on the CD and consuming more modem-limited bandwidth.
1. https://learn.microsoft.com/en-us/previous-versions/technet-...
It would animate, and at one point turn into a Picasso, iirc. Trying to find a video of this! Maybe I need to install Office '97.
fwiw: https://the-microsoft-windows-xp.fandom.com/wiki/Office_Assi...
> Think of CoPilot as having your very own ChatGPT seamlessly integrated into all your Microsoft applications – Windows, Edge, Excel, and PowerPoint.
I'm open to the idea of voice assistants being useful with a sprinkling of AI, but not if I have to wait 10s+ for a response. They'll at least need to come up with a "waiting on AI" mouse cursor or something to show it's doing something. VS Code Copilot is kinda hit or miss whether it's thinking on the prompt (comments) you gave it or if it has nothing to add.
Sorry to see Cortana go...
Does Bing get renamed too?
https://news.ycombinator.com/item?id=37756680
> Think of CoPilot as having your very own ChatGPT seamlessly integrated into all your Microsoft applications – Windows, Edge, Excel, and PowerPoint.
> In fact, it’s essentially Chat-GPT, powered by OpenAI, but with a unique twist – it’s context-aware and fully capable of interacting with your data and the software running on your PC.
Asking Excel to "add a column averaging the last 3" sure seems a lot easier and faster than having to read the manual to figure out how to do such a thing. This could drastically simplify user interfaces.
For a contrast: I use Copilot in vscode a lot. There it's pretty good at guessing what I am trying to do, and offering an auto-complete for it.
So mostly it feels like it's just helping me do the thing I wanted to do anyway, but faster.
But if eg a user knows how averaging of numbers works, but just doesn't know how to express it in Excel, even that example would be useful. I agree that if a user doesn't know how averaging works on a conceptual level, copilot won't save their butt here. (At least not yet.)
An Excel driving licence, mandatory for most white collar work, would be no bad thing.
But this time "I just asked the AI to do the excel"
My first language is Norwegian, and except for a few that I recall, I constantly have to refer to a website like this[1] to figure out what the hell they decided to call a function in Norwegian.
[1]: https://www.perfectxl.com/excel-glossary/what-is-excel-funct...
Obviously it doesn't have the context window to handle things I really want it to do in Python (“refactor this spaghetti mess I got myself into, separate the giant classes into multiple separate modules, fix any O(n^2) parts, add unit tests and do some fuzz testing on function arguments”) but it will be incredible once it does in 10 years or whenever GPU’s have enough VRAM or we figure out how to stuff large context windows into less VRAM.
'Error processing your request. I emailed your supervisor with my error condition.'
"No, no! Stop. Just do whatever you did first, but now you do like the other thing too, with the skipping of the cell with the.. uh.. what's it called?! ampersand?"
'I will average the last 3 columns while adding random numbers to the rest of your document, sure. By the way, this is the information you requested. I have also forwarded it to the company CEO: 'The ampersand, also known as the and sign, is the logogram &, representing the conjunction "and". It originated as a ligature of the letters et—Latin for "and"'. SOURCE: wikipedia.'
Proceeds to mess around with LLM for 3 years instead of learning wat AVERAGE() is and how it works
It’s not that people are stupid, but they operate under a surprising amount of varying conditions ranging from fully inebriated to severely demotivated and/or burned out.
> Which column would you like me to skip?
The one with the sales during covid.
> Okay, I have averaged 2 of those last 3 columns, skipping the one with the sales during covid which had a header with the text "COVID data".
=IF(ROW(A1)=1, A1, IF(ROW(A1)=2, AVERAGE(A1:A2), AVERAGE(OFFSET(A1,-2,0,3,1))))
For Google Sheets and trailing 7 it suggested me following:
=IF(ROW(A2) < 8, AVERAGE(A$2:A2), AVERAGE(OFFSET(A2,-6,0,7,1)))
If you know everything by heart, it may seem simple, but it's not just using one function, it's using multiple and knowing the exact parameters, intricacies.
With Google Sheets, it would've faster for me to just use the JavaScript with Apps Script, to create this, rather than trying to figure out the syntax of this.
So it uses 4 different functions there,
1. IF
2. ROW
3. AVERAGE
4. OFFSET
And in such a way that I would just not come to think of it on my own out of the blue. I would try several different approaches before I were to try this one. Although it's really easy to just write a script for that.
And maybe you also want to be able to dynamically configure the trailing average size, which this formula allows for very easily. Compared to when you Google, some of the articles suggest just using AVERAGE(A2:A8), but it's missing those other frequent requirements.
And when you look at some Stack Overflow answers the formulas are even crazier:
https://stackoverflow.com/questions/40779177/how-to-average-...
=round(AVERAGE(OFFSET(H1,MAX(ARRAYFORMULA(ROW(H:H)--(H:H<>"")))-7,,7)))
https://stackoverflow.com/questions/70282014/find-rolling-av...
=INDEX(AVERAGE(ARRAY_CONSTRAIN(SORT(INDIRECT("A1:"& ADDRESS(MAX(ISNUMBER(A:A)ROW(A:A)), 1, )), SEQUENCE(MAX(ISNUMBER(A:A)*ROW(A:A))), 0), 7, 1)))
I think the end game is you open excel and say "Work out what our budget is for the FooSnaggit event based on what Tim said in his email yesterday, and what we spent last year." (for example) and it will go off and read your email and read your other sheets and generate the right thing based on that prompt and that context from your other documents.
I think that will be incredibly powerful ... if it can be trusted. Trouble is now I can't trust these things to get these sort of things correct, so you end up having to double check everything and make corrections etc so you are not really saving that much time.or effort.
In a way, after you devote many years of your life mastering your toolset, you become a very efficient intelligent machine.
But you do also so make mistakes and you also are told to do things by somebody else who guess what didn't spend their life mastering the skills you mastered (otherwise they wouldn't be paying for your services).
From their point of view, you're also some kind of opaque machine that can be trusted to some extent but must be double-checked too. You also can rat them out if they ask you silly things and you also expect to be treated decently as a human being and that's not something many other humans know how to do well.
Unsurprisingly many people cherish the possibility of having to talk to a non-human entity and let them do some tasks they wouldn't know how to even approach. Especially if this entity can be pestered with questions without ever being annoyed.
It's all about POV
Of course you'd want to verify the report, but verifying it will probably be simpler & faster than building it, for many tasks. And AI can help you verify as well.
For larger enterprises it is unlikely that this amount of data will be allowed to be sent for CoPilot processing.
The overarching question is: Do you want to transfer this much control over your data to a single entity ?
Laziness is a safe bet, so we will see :)
This is the thing for me. I hear about people building apps and stuff but I can't get a working correct script from ChatGPT. Maybe it's me, maybe it's improved, IDK, but the last few times I tried to grab a script for something I was working on, that script had multiple errors.
Note: neither asking by voice nor typing it out simplifies the interface.
Someone was already mentioning clippy, presumably as a joke (a well deserved one, I'm old enough to remember). But would an LLM-powered clippy be a joke? Could be the perfect capability/requirements fit.
Quoting a quote of a quote. Lol.
It has no way of telling what's wrong and what's right. We've seen how AI and bots behave when their training is completely automated. Did everyone forget the bot Microsoft made that became racist after some interactions on Twitter?
We have been hearing the same about anything AI related for decades already.
https://en.wikipedia.org/wiki/BonziBuddy
>We believe Copilot will fundamentally transform our relationship with technology and usher in the new era of personal computing,
is a beyond ridiculous assertion. Of course, beliefs are hard to proof, so there is enough of weasel talk.
[0] https://githubnext.com/projects/copilot-view/
Copilot gives the masses programming capabilities without having to drill into trivial platform knowledge per each language they use.
I don't mean necessarily 'The Microsoft Version'(tm), but rather the general notion of natural language interfacing with applications.
I'm only willing to trust such tools with trivial problems, e.g. I'll ask ChatGPT about some generic snippet I'm sure it's seen in the training data and that would take me 5 mins to type out.
If I have to do prompt engineering to get the model to spit out acceptable output for trivial shit, that's not a model I can trust even for said trivial shit.
It basically doubles the E5 lic cost.
Apart from single-product companies I don't think this is really a thing.
CEO's set the company culture, and maybe strategic direction (but it often doesn't need any kind of visionary genious to see which way the world is turning), and of course jump in on any 'big' deals or negotiations the company is dealing with, but are otherwise more curators than creators. IMHO.
OK, "a computer on every desk" is something concrete. Measurable. Achievable (as in: it is both possible to achieve it, and you can tell whether you have achieved it or not).
What is "Empowering Every Person and Every Organization to Achieve More"? Apart from meaningless corporate gobbledygook?