Hi, I’m the author! The AI features were in beta for many months and I never got useful negative feedback on them. A couple people expressed displeasure about AI but I (mis)interpreted it as rage against the machine rather than actionable feedback. I think this was a mistake: every piece of beta feedback has a kernel of truth. Psychoanalyzing bug reports is a larger part of the job than I’d have expected when I began.
It was only after launch that I received genuinely thoughtful responses, such as how enterprise users would be unable to use software that includes this capability. I quit big tech a few years ago and maybe my sense for how big businesses work has become less finely honed. This is a blessing and a curse.
I guess the people that were outraged got what they wanted, but they’ve probably already jumped ship.
I hate AI being shoved into everything more than most, but honestly don’t think there was anything wrong with the original release (opt in, disabled by default).
I hate AI being shoved into everything more than most, but honestly don’t think there was anything wrong with the original release (opt in, disabled by default).
I tend to be skeptical of AI integrations into All The Things™ as well, I didn’t think there was a thing wrong with the original release either, and I don’t think this really adds anything from that perspective.
On the plus side, I think there is now a nicer worked example of a swift plugin for iTerm2 than there was before.
The degree of outrage was surprising; I sympathize with the maintainer here.
I’m not sure I would have taken too much notice of it. There is a ton of negative sentiment from tech people, particularly on the orange site. Best to ignore most of it and focus on people who have something constructive to offer.
I usually agree with hindsight being 20/20 but adding AI to a terminal application automatically on a minor software upgrade was not going be received well. Same would have applied to crypto…
It didn’t add it automatically. You had to complete several steps including providing your own API key for it to start working on your explicit request.
I am the first to object to random AI integration, have no use for this feature, and also have other reasons why I don’t and probably will never again use iTerm2. All of that said:
Although the feature was there in the core, it didn’t work out of the box. You needed to explicitly configure it to use an API key that you provided. I am quite surprised that it would be problematic for people.
It exists as a separate component to ensure that there’s no way to accidentally send information from the terminal over the network.
In what scenario would information accidentally be sent over the network? As a result of user error or programmer error? Both seem very unlikely. Justifications like this either appeal to or betray a fundamental misunderstanding of how programs work.
if I were the iTerm author and saw the lobster comments on the original release and then saw this comment after capitulating on the plugin I would throw iTerm in the garbage and go into the woods.
The author did engage with the community here when the feature was announced. I doubt lobste.rs was the only forum where people expressed concern.
I know this might be hard to understand or even be unfair, but so many companies have announced terrible “AI” features that the term has become almost radioactive. I believe the backlash against iTerm2 was so vehement because so many people instantly thought “omg it’s infected this respected application too!” It came from a place of disappointment, not hate.
Yes, exactly, the backlash was totally based on a feeling and not the reality of the specific situation.
So when the author moves the integration out to a plugin, to quell the not-entirely-rational panic, he gets told on here that he has “a fundamental misunderstanding of how programs work.”
EDIT: in other words, as an open source maintainer my takeaway from this community engagement would be “the peanut gallery will grind you down no matter what you do.”
It’s ambiguous. I read it as the author misunderstanding how programs work as well. In fact, I think it’s the more straightforward reading; “Justifications like this” - the author is the one making the justification (they literally are quoted), and so for the rest of the sentence it seems that the author is the one being referenced.
I made a general comment about a type of justification used in PR. In this particular instance, if you want to apply it to the maintainer of the open source project as the author of the web page (which is not certain to me) then clearly “the appealing to” case would be what applies here, as the maintainer is a well known experienced software developer.
Second, to claim that a statement “betrays” an attribute is not identical to claiming that the person in question has that attribute. There is a useful nuance here that separates one’s identity from what their actions may intentionally or unintentionally indicate.
If you bought an API key by accident, entered it into iTerm2’s settings by accident, and invoked the feature without meaning to, information could have been sent to OpenAI.
Hi, I’m the author! The AI features were in beta for many months and I never got useful negative feedback on them. A couple people expressed displeasure about AI but I (mis)interpreted it as rage against the machine rather than actionable feedback. I think this was a mistake: every piece of beta feedback has a kernel of truth. Psychoanalyzing bug reports is a larger part of the job than I’d have expected when I began.
It was only after launch that I received genuinely thoughtful responses, such as how enterprise users would be unable to use software that includes this capability. I quit big tech a few years ago and maybe my sense for how big businesses work has become less finely honed. This is a blessing and a curse.
Thanks for all the blood and sweat you’ve put into iTerm, George.
Thank you for iTerm!
Thanks for the time & effort you’ve put into iTerm!
I guess the people that were outraged got what they wanted, but they’ve probably already jumped ship.
I hate AI being shoved into everything more than most, but honestly don’t think there was anything wrong with the original release (opt in, disabled by default).
I tend to be skeptical of AI integrations into All The Things™ as well, I didn’t think there was a thing wrong with the original release either, and I don’t think this really adds anything from that perspective.
On the plus side, I think there is now a nicer worked example of a swift plugin for iTerm2 than there was before.
It boggles my mind they didn’t start with this plugin/add-on implementation but I’m glad to see they listened to their users.
Everything always looks obvious in hindsight though.
The degree of outrage was surprising; I sympathize with the maintainer here.
I’m surprised that you find it surprising, genuinely. Every hype cycle creates a corresponding backlash.
I’m not sure I would have taken too much notice of it. There is a ton of negative sentiment from tech people, particularly on the orange site. Best to ignore most of it and focus on people who have something constructive to offer.
I usually agree with hindsight being 20/20 but adding AI to a terminal application automatically on a minor software upgrade was not going be received well. Same would have applied to crypto…
It didn’t add it automatically. You had to complete several steps including providing your own API key for it to start working on your explicit request.
I am the first to object to random AI integration, have no use for this feature, and also have other reasons why I don’t and probably will never again use iTerm2. All of that said:
Although the feature was there in the core, it didn’t work out of the box. You needed to explicitly configure it to use an API key that you provided. I am quite surprised that it would be problematic for people.
In what scenario would information accidentally be sent over the network? As a result of user error or programmer error? Both seem very unlikely. Justifications like this either appeal to or betray a fundamental misunderstanding of how programs work.
It’s happened before in this very project: https://www.bleepingcomputer.com/news/security/iterm2-leaks-everything-you-hover-in-your-terminal-via-dns-requests/
That feature was mandatory when it was initially released. The AI feature was opt-in when it was released. My question remains.
if I were the iTerm author and saw the lobster comments on the original release and then saw this comment after capitulating on the plugin I would throw iTerm in the garbage and go into the woods.
The author did engage with the community here when the feature was announced. I doubt lobste.rs was the only forum where people expressed concern.
I know this might be hard to understand or even be unfair, but so many companies have announced terrible “AI” features that the term has become almost radioactive. I believe the backlash against iTerm2 was so vehement because so many people instantly thought “omg it’s infected this respected application too!” It came from a place of disappointment, not hate.
Yes, exactly, the backlash was totally based on a feeling and not the reality of the specific situation.
So when the author moves the integration out to a plugin, to quell the not-entirely-rational panic, he gets told on here that he has “a fundamental misunderstanding of how programs work.”
EDIT: in other words, as an open source maintainer my takeaway from this community engagement would be “the peanut gallery will grind you down no matter what you do.”
I don’t think you’ll find in this thread any claim made about the maintainer.
You said:
It’s addressing users who complained, not the maintainer
It’s ambiguous. I read it as the author misunderstanding how programs work as well. In fact, I think it’s the more straightforward reading; “Justifications like this” - the author is the one making the justification (they literally are quoted), and so for the rest of the sentence it seems that the author is the one being referenced.
I made a general comment about a type of justification used in PR. In this particular instance, if you want to apply it to the maintainer of the open source project as the author of the web page (which is not certain to me) then clearly “the appealing to” case would be what applies here, as the maintainer is a well known experienced software developer.
Second, to claim that a statement “betrays” an attribute is not identical to claiming that the person in question has that attribute. There is a useful nuance here that separates one’s identity from what their actions may intentionally or unintentionally indicate.
I’m just explaining why I read it that way.
If you bought an API key by accident, entered it into iTerm2’s settings by accident, and invoked the feature without meaning to, information could have been sent to OpenAI.