I'm really worried about what happens when we mix these kind of dependency-oriented business strategies with the sensitivity of a human's personal context. With no "file format" equivalent for personal context, how do I ever switch "assistant" in this provisional future?
Microsoft can't learn a thing from their own history. Perhaps if they made a product that was useful and not deceiving their users they'd have more success. It seems they aren't capable of that anymore.
I'd really be worried if a competent company was trying to make people addicted. This is absolutely a thing. Social media, mobile games, LLMs, tobacco.
Making a useful product is far more difficult than tapping into our base desires as human beings. Microsoft hasn't been an innovative company in a long time and that's by design.
That’s always been Microsoft in a nutshell though. They constantly stumble around never making anything truly great but still doing a good enough job to keep their existing customers.
Scout sounds like an excitable little dog that runs headlong into trees when trying to catch a frisbee.
Given Microsoft's long history of failure with personal assistants I'm looking forward to this one! Clippy, Cortana, Copilot! Wasn't an animated dog called rover one of these way back? The best of all was unquestionably Ms. Dewey for Microsoft Windows Live Search who is almost forgotten.
Nadella is good at messaging. It is difficult to actually believe he wants to do anything but push it as far and wide as possible, and addiction is fine. It's just spin on the wording.
I've met plenty of people who want to make products that solve problems, even if the product's user only has those problems once in a while. Reaching for a well-liked, well-matched tool whenever a problem arises isn't addicted or quasi-addicted or "as though" addicted behavior.
Once you're thinking about how to keep a user coming back, you're in the mutually adversarial design space, whatever language is used to more pleasantly redecorate that reality.
You can't be a good designer if you aren't thinking about how to get your users to love your product so much that they keep coming back. There are good and bad ways to keep users coming back. The good way is to simply make the product very useful. The bad way is to make the user psychologically dependent on your product in some way.
Yet almost everyone uses dark patterns, which imply they don't think their product is good enough for users to return on their own volition. In fact, I can't think of a single for-profit company that doesn't use at least one dark pattern.
> Anyone who makes products want users of our product to keep coming back as though they are addicted, but not actually addicted.
Can you explain the distinction? I am not seeing it. If I keep refreshing a product page to get another dopamine hit, am I addicted or not addicted but appearing so to your metrics?
Everyone likes a beer analogy (almost as much as CS teachers love car analogies!) so I’ll try and do one that applies in the way I _think_ GP intends:
Brewers want people to want beer, and to perhaps puritans, that desire could appear as “addicted”. However, brewers don’t want addicts - liver failure, destitution, death, are all things I doubt a brewer wants to see in their consumer base because you can’t drink if you don’t have a liver, don’t have money, or don’t have life.
Did I, as a child, think my dad was addicted to alcohol because I saw him drink everyday? I did, that’s the appearance it gave. Was he? Not to the clinical point of addiction, technically - he functioned, maintained relationships and a job, and wasn’t more than occasionally emotionally abusive. He fit the type of customer GP seems to talk about - appearing to be addicted but not wholly, truly addicted.
Are you addicted to your job? You keep going back every single work day. Does that mean you are addicted? Just because you keep repeating an action doesn't mean you're addicted. It just means it is a solving a problem for you (such as providing you with a salary to buy food and pay rent) and does it well.
I am not addicted to my job but my employer would like me to be.
I think apps are a different beast. They (generally, with few exceptions) want their users to be addicted. An addicted user is more likely to come back than one that gets a need met. Once that need is fulfilled, they leave.
If companies actually wanted to fill people's needs they wouldn't use dark patterns like having to call to cancel, spamming them without their consent, switching opt-out choices back with updates, etc. Because they use these dirty tricks, it's hard to believe they have the users best interest in mind. They don't. They just want the line to go up.
There's a race and tug-of-war to frame how interaction with apps works. The addiction word has a strong "think of the children" energy and I would expand any company to want to have their app tagged with the term.
Of course, what exactly "addicted" means in the context of interacting with a program really pretty fuzzy but yeah, "users not in control of themselves" is perhaps the biggest implication (and not necessarily false, mind you). Of course, this is a matter of both degree and social context.
If only we had a social dialog about the real meaning of things labeled addictive, perhaps their terrible impact could be mitigated. But hey, I guess we get policing and moral panics instead.
>Microsoft has been piloting Scout as an internal tool for employees it was calling “ClawPilot,” since March. ClawPilot—and now Scout—are part of “Project Lobster,” which is a Microsoft plan to bring the popular OpenClaw AI tool to its Microsoft 365 suite of products in a way that nontechnical people can use.
It is "addictive" in the sense that it works really well, and has some guardrails so the risk of it doing something insane is minimized. I have done some cool stuff with it!
In an otherwise pleasant, humanist framing, they jarringly conclude Microsoft's primary AI application will be putting people into parasocial AI relationships for profit.
I'm confused and dissapointed that this isn't called Copilot, the users want more things to be called copilot even if they aren't related to each other, consider renaming Scout to Copilot, or at least Scout Copilot, or even better Copilot* (*Copilot Scout)
I'm really worried about what happens when we mix these kind of dependency-oriented business strategies with the sensitivity of a human's personal context. With no "file format" equivalent for personal context, how do I ever switch "assistant" in this provisional future?
Microsoft can't learn a thing from their own history. Perhaps if they made a product that was useful and not deceiving their users they'd have more success. It seems they aren't capable of that anymore.
They're those guys that spend hundreds of hours in pickup artist courses rather than being a person people would like to be with, yeah.
They are also just bad at almost everything they do. I'd really be worried if a competent company was trying to make people addicted.
Although maybe it doesn't take much, given that it sounded at one point like the Microsoft execs were addicted already.
I'd really be worried if a competent company was trying to make people addicted. This is absolutely a thing. Social media, mobile games, LLMs, tobacco.
> They are also just bad at almost everything they do
Huh? Have you forgotten Clippy, the first AI agent?
/s
Making a useful product is far more difficult than tapping into our base desires as human beings. Microsoft hasn't been an innovative company in a long time and that's by design.
That’s always been Microsoft in a nutshell though. They constantly stumble around never making anything truly great but still doing a good enough job to keep their existing customers.
Scout sounds like an excitable little dog that runs headlong into trees when trying to catch a frisbee.
Given Microsoft's long history of failure with personal assistants I'm looking forward to this one! Clippy, Cortana, Copilot! Wasn't an animated dog called rover one of these way back? The best of all was unquestionably Ms. Dewey for Microsoft Windows Live Search who is almost forgotten.
https://en.wikipedia.org/wiki/Microsoft_Bob
You can go straight to hell for making me feel old.
As a kid I absolutely loved these interfaces.
it was a knockoff of https://en.wikipedia.org/wiki/Magic_Cap
I still do, honestly.
guess who is reusing that name?
https://bob.ibm.com/
I saw IBMers wearing shirts with it at the moscone center this week at the snowflake event.
Don't forget about Tay https://en.wikipedia.org/wiki/Tay_(chatbot)
I thought his name was Watson (the dog assistant).
Edit: I was wrong. But there is "Power Pup" and apparently Will Shakespeare
Power pup was the superdog. Rocky was the normal dog.
(I was way too into the building apps with the assistant framework as a kid...)
It was Rover
Statement from Nadella: https://www.404media.co/satya-nadella-not-sure-who-said-micr...
Nadella is good at messaging. It is difficult to actually believe he wants to do anything but push it as far and wide as possible, and addiction is fine. It's just spin on the wording.
> Microsoft's CEO seems unaware of what's going on at his own company.
damn, he quickly disavowed the statement and got panned anyway
would falling on the sword worked better??
Being aware of what the basic strategy is with one of their most high profile products would have worked better.
People keep telling me it’s why we even have C-level folks at the top!
He's lying or incompetent, your choice
Both of those things can be true at the same time for some people.
It's kind of sad that they didn't call this Cortana. I guess they made that name too toxic?
"Addiction" is a bad word. It implies the user is not in control of themselves.
Anyone who makes products want users of our product to keep coming back as though they are addicted, but not actually addicted.
I've met plenty of people who want to make products that solve problems, even if the product's user only has those problems once in a while. Reaching for a well-liked, well-matched tool whenever a problem arises isn't addicted or quasi-addicted or "as though" addicted behavior.
Once you're thinking about how to keep a user coming back, you're in the mutually adversarial design space, whatever language is used to more pleasantly redecorate that reality.
You can't be a good designer if you aren't thinking about how to get your users to love your product so much that they keep coming back. There are good and bad ways to keep users coming back. The good way is to simply make the product very useful. The bad way is to make the user psychologically dependent on your product in some way.
Yet almost everyone uses dark patterns, which imply they don't think their product is good enough for users to return on their own volition. In fact, I can't think of a single for-profit company that doesn't use at least one dark pattern.
> Anyone who makes products want users of our product to keep coming back as though they are addicted, but not actually addicted.
Can you explain the distinction? I am not seeing it. If I keep refreshing a product page to get another dopamine hit, am I addicted or not addicted but appearing so to your metrics?
Everyone likes a beer analogy (almost as much as CS teachers love car analogies!) so I’ll try and do one that applies in the way I _think_ GP intends:
Brewers want people to want beer, and to perhaps puritans, that desire could appear as “addicted”. However, brewers don’t want addicts - liver failure, destitution, death, are all things I doubt a brewer wants to see in their consumer base because you can’t drink if you don’t have a liver, don’t have money, or don’t have life.
Did I, as a child, think my dad was addicted to alcohol because I saw him drink everyday? I did, that’s the appearance it gave. Was he? Not to the clinical point of addiction, technically - he functioned, maintained relationships and a job, and wasn’t more than occasionally emotionally abusive. He fit the type of customer GP seems to talk about - appearing to be addicted but not wholly, truly addicted.
I think the point the gp is making that companies want their users addicted but never should say "addicted" since it has undesirable implications.
Are you addicted to your job? You keep going back every single work day. Does that mean you are addicted? Just because you keep repeating an action doesn't mean you're addicted. It just means it is a solving a problem for you (such as providing you with a salary to buy food and pay rent) and does it well.
I am not addicted to my job but my employer would like me to be.
I think apps are a different beast. They (generally, with few exceptions) want their users to be addicted. An addicted user is more likely to come back than one that gets a need met. Once that need is fulfilled, they leave.
If companies actually wanted to fill people's needs they wouldn't use dark patterns like having to call to cancel, spamming them without their consent, switching opt-out choices back with updates, etc. Because they use these dirty tricks, it's hard to believe they have the users best interest in mind. They don't. They just want the line to go up.
why do you think microsoft is that concerned with their user's wellbeing?
there are other industries who's entire business revovles around selling to addicts, why would MS of all companies suddenly balk at that line?
Indeed,
There's a race and tug-of-war to frame how interaction with apps works. The addiction word has a strong "think of the children" energy and I would expand any company to want to have their app tagged with the term.
Of course, what exactly "addicted" means in the context of interacting with a program really pretty fuzzy but yeah, "users not in control of themselves" is perhaps the biggest implication (and not necessarily false, mind you). Of course, this is a matter of both degree and social context.
If only we had a social dialog about the real meaning of things labeled addictive, perhaps their terrible impact could be mitigated. But hey, I guess we get policing and moral panics instead.
What, a Microsoft AI product that isn't branded "Copilot"?!
From the article this one links to:
>Microsoft has been piloting Scout as an internal tool for employees it was calling “ClawPilot,” since March. ClawPilot—and now Scout—are part of “Project Lobster,” which is a Microsoft plan to bring the popular OpenClaw AI tool to its Microsoft 365 suite of products in a way that nontechnical people can use.
Lately they've also been talking about autopilots.
Clippy would have been perfect for the role!
Discussion (59 points, 3 days ago, 65 comments) https://news.ycombinator.com/item?id=48374503
Would be worrisome if they were good at building stuff.
I have been using it for some weeks now.
It is "addictive" in the sense that it works really well, and has some guardrails so the risk of it doing something insane is minimized. I have done some cool stuff with it!
This helps explain the dissonance in Microsoft's recent Humanist Superintelligence article about creating "AI companions for everyone".
https://microsoft.ai/news/towards-humanist-superintelligence...
In an otherwise pleasant, humanist framing, they jarringly conclude Microsoft's primary AI application will be putting people into parasocial AI relationships for profit.
Feedback to microsoft:
I'm confused and dissapointed that this isn't called Copilot, the users want more things to be called copilot even if they aren't related to each other, consider renaming Scout to Copilot, or at least Scout Copilot, or even better Copilot* (*Copilot Scout)
It's technically called Autopilot* Scout if you can believe it...
I think Teams should be renamed Scout Messaging, and Outlook to Scout Messaging: Legacy.
Co-Pilot is a perfect name because it, just like actual Co-Pilots, perform their best when they do nothing