Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity
-
Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity
"When developers are allowed to use AI tools, they take 19% longer to complete issues—a significant slowdown that goes against developer beliefs and expert forecasts. This gap between perception and reality is striking: developers expected AI to speed them up by 24%, and even after experiencing the slowdown, they still believed AI had sped them up by 20%."
Like so many papers on AI, this ignores the ethical aspects and any power analysis -- which would be just as problematic if AI tools actually made experienced developers faster. So it's the gap between perception and reality that I find most interesting. Keep that in mind when you see developers -- or anybody else -- talking about how amazingly more productive they are thanks to AI!
-
Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity
"When developers are allowed to use AI tools, they take 19% longer to complete issues—a significant slowdown that goes against developer beliefs and expert forecasts. This gap between perception and reality is striking: developers expected AI to speed them up by 24%, and even after experiencing the slowdown, they still believed AI had sped them up by 20%."
Like so many papers on AI, this ignores the ethical aspects and any power analysis -- which would be just as problematic if AI tools actually made experienced developers faster. So it's the gap between perception and reality that I find most interesting. Keep that in mind when you see developers -- or anybody else -- talking about how amazingly more productive they are thanks to AI!
@thenexusofprivacy this is still a useful argument to those who don’t care about ethics, like some corporations.
-
Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity
"When developers are allowed to use AI tools, they take 19% longer to complete issues—a significant slowdown that goes against developer beliefs and expert forecasts. This gap between perception and reality is striking: developers expected AI to speed them up by 24%, and even after experiencing the slowdown, they still believed AI had sped them up by 20%."
Like so many papers on AI, this ignores the ethical aspects and any power analysis -- which would be just as problematic if AI tools actually made experienced developers faster. So it's the gap between perception and reality that I find most interesting. Keep that in mind when you see developers -- or anybody else -- talking about how amazingly more productive they are thanks to AI!
Bluesky is starting age verification in the UK on July 25
"We’ll use Epic Games’ Kids Web Services (KWS) to give our UK community choices about how to verify their age. If you’re in the UK, you can choose between methods like payment card verification, ID scans, and face scans. (See here to learn more about how KWS safeguards user information.) For people who are under 18 or don’t want to go through this process, we’ll make adult-appropriate content inaccessible, and we’ll disable features like direct messaging."
"Highly effective age assurance” is a requirement of the Online Safety Act, so they don't really have a lot of choice. Still, not good! And with the US Supreme Court recently upholding Texas' age verification law, we should expect to see similar policies in other geographies.
-
Bluesky is starting age verification in the UK on July 25
"We’ll use Epic Games’ Kids Web Services (KWS) to give our UK community choices about how to verify their age. If you’re in the UK, you can choose between methods like payment card verification, ID scans, and face scans. (See here to learn more about how KWS safeguards user information.) For people who are under 18 or don’t want to go through this process, we’ll make adult-appropriate content inaccessible, and we’ll disable features like direct messaging."
"Highly effective age assurance” is a requirement of the Online Safety Act, so they don't really have a lot of choice. Still, not good! And with the US Supreme Court recently upholding Texas' age verification law, we should expect to see similar policies in other geographies.
If you're wondering about how the Online Safety Act applies here in the fediverse ... it's complicated.
@iftas has a lot of resources at https://connect.iftas.org/library/legal-regulatory/online-safety-act/
-
If you're wondering about how the Online Safety Act applies here in the fediverse ... it's complicated.
@iftas has a lot of resources at https://connect.iftas.org/library/legal-regulatory/online-safety-act/
@thenexusofprivacy @iftas I'm anticipating basically just needing to be on a permanent VPN connection in the near-ish future.
-
@thenexusofprivacy this is still a useful argument to those who don’t care about ethics, like some corporations.
@alwayscurious Agreed! That said, people who want to deploy AI will point out that tools are getting better, and come up with reasons that their situation is different than the specific one discussed in the study ... so tbd how useful it winds up being in practice on that front.
-
If you're wondering about how the Online Safety Act applies here in the fediverse ... it's complicated.
@iftas has a lot of resources at https://connect.iftas.org/library/legal-regulatory/online-safety-act/
And here's a timely article by @mgeist about The Legal and Privacy Concerns of Mandatory Age Verification Technologies. It focus on Canadian legislation, but the risks and harms it discusses are applicable everywhere.
"A harm that proponents of these technologies want to mitigate is “the extent of youths’ exposure to sexually explicit material online, the frequency with which this material is of an aggressive or violent nature, and the potential harms to body image or mental health it may cause”. Opponents note that these technologies are harmful in that they would limit young people’s (especially those from marginalized groups) “access to online content or forums” that provide “avenues for community-building, civic engagement, and education”, as well as “self-discovery”. Further, use of such technologies comes with risk of data breaches that would publicly expose people’s online activities, consequently causing “psychological or physical harms” and possibly discouraging them from “operating freely in the digital environment”"
-
@thenexusofprivacy @iftas I'm anticipating basically just needing to be on a permanent VPN connection in the near-ish future.
@ret @thenexusofprivacy @iftas but what if i don't have a VPN? and what if i don't want to do face scanning or payment card verification? i'd much rather just exist on the internet as a 21 year old adult without any of this OSA nonsense.
-
@ret @thenexusofprivacy @iftas but what if i don't have a VPN? and what if i don't want to do face scanning or payment card verification? i'd much rather just exist on the internet as a 21 year old adult without any of this OSA nonsense.
@cardboard @thenexusofprivacy @iftas I guess none of us are going to have a choice in that. Either have your face or ID scanned or don’t use the services.
-
@cardboard @thenexusofprivacy @iftas I guess none of us are going to have a choice in that. Either have your face or ID scanned or don’t use the services.
@ret @thenexusofprivacy @iftas so does that mean i have to give up discord, telegram and here, among other things?
-
@ret @thenexusofprivacy @iftas so does that mean i have to give up discord, telegram and here, among other things?
@cardboard @thenexusofprivacy @iftas potentially? I am no oracle, but it does seem like at least the centralised ones will start cracking down soon.
-
@cardboard @thenexusofprivacy @iftas potentially? I am no oracle, but it does seem like at least the centralised ones will start cracking down soon.
Yeah, it's hard to know but the trend is certainly in that direction. It's possible that small fedi instances without open registration might not be subject to this for the Online Safety Act (although that's certainly not legal advice!), but the EU is working on an age verification requirement, so is Canada, and there are bills in the US that could require it as well ... so it's hard to be optimistic.
-
@ret @thenexusofprivacy @iftas so does that mean i have to give up discord, telegram and here, among other things?
@cardboard @ret @thenexusofprivacy @iftas Do want to point out it seems you can still use the sites,
"If users don’t want to verify their age, or if they’re under 18, they will still be able to have an account with certain features limited. Bluesky will block “adult-appropriate content” and turn off certain features, such as direct messaging."
-
@cardboard @ret @thenexusofprivacy @iftas Do want to point out it seems you can still use the sites,
"If users don’t want to verify their age, or if they’re under 18, they will still be able to have an account with certain features limited. Bluesky will block “adult-appropriate content” and turn off certain features, such as direct messaging."
True, but what is "adult-approriate content"? From what I understand about the dynamics in the UK, that could well include anything LGBTQIA2S+ related -- here in the US, many of the supporters of the proposed KOSA law (which would in practice require age verification) make no secret of the fact that it's targeting LGBTIA2S+ people. And things could certainly go that route for other "controversial" topics ...we'll see how things play out, but the scope is likely to steadily increase over time.
-
True, but what is "adult-approriate content"? From what I understand about the dynamics in the UK, that could well include anything LGBTQIA2S+ related -- here in the US, many of the supporters of the proposed KOSA law (which would in practice require age verification) make no secret of the fact that it's targeting LGBTIA2S+ people. And things could certainly go that route for other "controversial" topics ...we'll see how things play out, but the scope is likely to steadily increase over time.
And this just in via @dell here in the US, DHS' threat bulletins are describing ordinary behavior, like riding a skateboard or livestreaming, as threatening to law enforcement. OMG, we certainly wouldn't want to expose kids to that kind of threatening behavior!!!!!!!
https://www.wired.com/story/dhs-tells-police-that-common-protest-activities-are-violent-tactics/
-
T thenexusofprivacy@infosec.exchange shared this topic
-
True, but what is "adult-approriate content"? From what I understand about the dynamics in the UK, that could well include anything LGBTQIA2S+ related -- here in the US, many of the supporters of the proposed KOSA law (which would in practice require age verification) make no secret of the fact that it's targeting LGBTIA2S+ people. And things could certainly go that route for other "controversial" topics ...we'll see how things play out, but the scope is likely to steadily increase over time.
@thenexusofprivacy @cardboard @ret Yeah tho I don't see this law lasting long, its already falling apart.
-
@thenexusofprivacy @cardboard @ret Yeah tho I don't see this law lasting long, its already falling apart.
Maybe ... although it could go the other way too -- early this year I saw somebody in the UK Government saying the Online Safety Act needed to be strengthened .
There's a "repeal" petition here with a few thousand votes but I don't know if there's any organized campaign yet. https://petition.parliament.uk/petitions/722903
-
Maybe ... although it could go the other way too -- early this year I saw somebody in the UK Government saying the Online Safety Act needed to be strengthened .
There's a "repeal" petition here with a few thousand votes but I don't know if there's any organized campaign yet. https://petition.parliament.uk/petitions/722903
@thenexusofprivacy @TheVampireFishQueen @cardboard signed... for what it's worth. I commented on the calls for the act to be expanded today too: https://furry.engineer/@ret/114835043899065605
-
@thenexusofprivacy @TheVampireFishQueen @cardboard signed... for what it's worth. I commented on the calls for the act to be expanded today too: https://furry.engineer/@ret/114835043899065605
Exactly. The original proposal for the Online Safety Act included regulating this "harmful but not illegal" content, but it got stripped out (it was so bad that even Kemi Badenoch was against it). Still, the govenment wants more power and "this will make everyody safer" is often a good way to spin it.