Erica Kangas
@ericaashley.com
👩🏻🌾 Tech • Data • Social Impact
🗺️ Boston, MA
👩🏻💻 Head of Engineering @ Scroobious • Mentor, building the next generation of Boston tech talent
🗺️ Boston, MA
👩🏻💻 Head of Engineering @ Scroobious • Mentor, building the next generation of Boston tech talent
Reposted by Erica Kangas
Reposted by Erica Kangas
Thoughtful people are capable of asking themselves:
- do I think gatekeeping is the path to shared safety?
- do I think x degree is the only way to select for y behavior?
- is there a reason I think Being Technical never has to be explained or defined?
- who is the one who gets to enforce it, then?
- do I think gatekeeping is the path to shared safety?
- do I think x degree is the only way to select for y behavior?
- is there a reason I think Being Technical never has to be explained or defined?
- who is the one who gets to enforce it, then?
November 16, 2025 at 5:14 PM
Thoughtful people are capable of asking themselves:
- do I think gatekeeping is the path to shared safety?
- do I think x degree is the only way to select for y behavior?
- is there a reason I think Being Technical never has to be explained or defined?
- who is the one who gets to enforce it, then?
- do I think gatekeeping is the path to shared safety?
- do I think x degree is the only way to select for y behavior?
- is there a reason I think Being Technical never has to be explained or defined?
- who is the one who gets to enforce it, then?
Reposted by Erica Kangas
An AI pioneer is convinced that most in his field have been led astray by the siren song of large language models.
He’s Been Right About AI for 40 Years. Now He Thinks Everyone Is Wrong.
Yann LeCun invented many fundamental components of modern AI. Now he’s convinced most in his field have been led astray by the siren song of large language models.
on.wsj.com
November 15, 2025 at 11:10 AM
An AI pioneer is convinced that most in his field have been led astray by the siren song of large language models.
Reposted by Erica Kangas
A growing number of SEC filings now list AI as a material risk, reflecting both its rapid adoption and the uncertainty surrounding how to control it.
The new AI warnings popping up in SEC filings
A growing number of SEC filings now list AI as a material risk, reflecting both its rapid adoption and the uncertainty surrounding how to control it.
www.businessinsider.com
November 13, 2025 at 12:40 PM
A growing number of SEC filings now list AI as a material risk, reflecting both its rapid adoption and the uncertainty surrounding how to control it.
Reposted by Erica Kangas
I have a feature essay for The Guardian today on the mirage of AI medicine, why care cannot be automated, and how overwhelming uptake of AI by American health capitalism threatens to undermine the very possibility of democracy.
www.theguardian.com/us-news/ng-i...
www.theguardian.com/us-news/ng-i...
What we lose when we surrender care to algorithms | Eric Reinhart
A dangerous faith in AI is sweeping American healthcare – with consequences for the basis of society itself
www.theguardian.com
November 9, 2025 at 4:05 PM
I have a feature essay for The Guardian today on the mirage of AI medicine, why care cannot be automated, and how overwhelming uptake of AI by American health capitalism threatens to undermine the very possibility of democracy.
www.theguardian.com/us-news/ng-i...
www.theguardian.com/us-news/ng-i...
Reposted by Erica Kangas
In @nytopinion.nytimes.com
“If Americans want to challenge their country’s illiberal turn, they need to stop clinging to the recent past,” Sven Beckert says in a guest essay. “Like other economic regimes before it, it is gone. Resurrection is impossible and to aim for it is politically disastrous.”
“If Americans want to challenge their country’s illiberal turn, they need to stop clinging to the recent past,” Sven Beckert says in a guest essay. “Like other economic regimes before it, it is gone. Resurrection is impossible and to aim for it is politically disastrous.”
Opinion | Capitalism Is Constantly Reinventing Itself. That’s Why Things Feel So Volatile.
We are in the middle of the capitalistic order reinventing itself.
nyti.ms
November 5, 2025 at 4:15 AM
In @nytopinion.nytimes.com
“If Americans want to challenge their country’s illiberal turn, they need to stop clinging to the recent past,” Sven Beckert says in a guest essay. “Like other economic regimes before it, it is gone. Resurrection is impossible and to aim for it is politically disastrous.”
“If Americans want to challenge their country’s illiberal turn, they need to stop clinging to the recent past,” Sven Beckert says in a guest essay. “Like other economic regimes before it, it is gone. Resurrection is impossible and to aim for it is politically disastrous.”
Reposted by Erica Kangas
Sound analysis from economist Muhamad Chatib Basri on Indonesia's good economic growth but fragile middle class. He helps to describe why millions of young Indonesian were involved in street protests nationwide
carnegieendowment.org/research/202...
carnegieendowment.org/research/202...
November 4, 2025 at 9:12 AM
Sound analysis from economist Muhamad Chatib Basri on Indonesia's good economic growth but fragile middle class. He helps to describe why millions of young Indonesian were involved in street protests nationwide
carnegieendowment.org/research/202...
carnegieendowment.org/research/202...
Reposted by Erica Kangas
Can you get promoted with "mentors juniors" as a component of that promotion? No....? Then we've really done something specific as a design choice for the teams there
November 3, 2025 at 1:15 AM
Can you get promoted with "mentors juniors" as a component of that promotion? No....? Then we've really done something specific as a design choice for the teams there
It’s the economy, stupid
From Amazon to Target, the country’s largest employers are retrenching, laying off tens of thousands of office workers.
Tens of Thousands of White-Collar Jobs Are Disappearing as AI Starts to Bite
Companies ranging from Amazon to Target are sending workers into an unwelcoming market.
on.wsj.com
October 29, 2025 at 9:54 PM
It’s the economy, stupid
Reposted by Erica Kangas
The end result of the GOP's census campaign could not only radically alter the type of data made available, but also put the personal data of every person living in the US at risk.
Republican plan would make deanonymization of census data trivial
“Differential privacy” algorithm prevents statistical data from being tied to individuals.
arstechnica.com
October 29, 2025 at 5:05 PM
The end result of the GOP's census campaign could not only radically alter the type of data made available, but also put the personal data of every person living in the US at risk.
Reposted by Erica Kangas
Reposted by Erica Kangas
And I've shared it before but our report on the likely harms of AI in education singled out the risks of genAI being used to make decisions or judgements about students that nobody can fully explain - or counter. Yet this is what edtech vendors are now pushing hard nepc.colorado.edu/publication/ai
Time for a Pause: Without Effective Public Oversight, AI in Schools Will Do More Harm Than Good.
Ignoring their own well-publicized calls to regulate AI development and to pause implementation of its applications, major technology companies such as Google, Microsoft, and Meta are racing to fend o...
nepc.colorado.edu
October 21, 2025 at 11:03 PM
And I've shared it before but our report on the likely harms of AI in education singled out the risks of genAI being used to make decisions or judgements about students that nobody can fully explain - or counter. Yet this is what edtech vendors are now pushing hard nepc.colorado.edu/publication/ai
Reposted by Erica Kangas
*puts my contrarian hat on*
not having code is also a liability!
Lots of times people want to solve social problems with tech. Don’t do that. But sometimes (OCCASIONALLY) people are desperately trying to solve tech problems with people in order to avoid writing code
That’s not great either!
not having code is also a liability!
Lots of times people want to solve social problems with tech. Don’t do that. But sometimes (OCCASIONALLY) people are desperately trying to solve tech problems with people in order to avoid writing code
That’s not great either!
yessss, grasshopper. code is liability. carry on 📈✨
my toxic engineer trait is that I'm basically allergic to code
all code is bad. the less code the better. why are you writing new code stahp *sprays devs with squirt bottle*
all code is bad. the less code the better. why are you writing new code stahp *sprays devs with squirt bottle*
October 21, 2025 at 12:50 AM
*puts my contrarian hat on*
not having code is also a liability!
Lots of times people want to solve social problems with tech. Don’t do that. But sometimes (OCCASIONALLY) people are desperately trying to solve tech problems with people in order to avoid writing code
That’s not great either!
not having code is also a liability!
Lots of times people want to solve social problems with tech. Don’t do that. But sometimes (OCCASIONALLY) people are desperately trying to solve tech problems with people in order to avoid writing code
That’s not great either!
Reposted by Erica Kangas
Also if the ultimate test of "are you a good dev" isn't how fast you work, why should the ultimate test of "is this a tool that helps people do good dev" be "it makes you faster"......?
October 12, 2025 at 10:37 AM
Also if the ultimate test of "are you a good dev" isn't how fast you work, why should the ultimate test of "is this a tool that helps people do good dev" be "it makes you faster"......?
Reposted by Erica Kangas
Can't believe this happened in Massachusetts. There are still some things in this world where accuracy is important 🙄
NEW: The use of AI to score Massachusetts’ standardized MCAS tests resulted in errors that affected hundreds of exams, the NBC10 Investigators learned. on.nbcboston.com/FGDbuPo
'No rhyme or reason': AI grading issue affects hundreds of MCAS essays
The use of AI to score Massachusetts’ standardized MCAS tests resulted in errors that affected hundreds of exams, the NBC10 Investigators learned.
on.nbcboston.com
September 20, 2025 at 12:49 PM
Can't believe this happened in Massachusetts. There are still some things in this world where accuracy is important 🙄
Reposted by Erica Kangas
So, Generative AI is a stochastic parrot regurgitating the average of all the world’s most mediocre code… Yet, somehow, the “upper elite 1% of talent” is getting massive productivity gains from it? To the point where Elite Performing tech corporations mandate it?
That doesn’t add up
That doesn’t add up
September 18, 2025 at 9:19 PM
So, Generative AI is a stochastic parrot regurgitating the average of all the world’s most mediocre code… Yet, somehow, the “upper elite 1% of talent” is getting massive productivity gains from it? To the point where Elite Performing tech corporations mandate it?
That doesn’t add up
That doesn’t add up
Reposted by Erica Kangas
Shocked but not surprised to learn how many ~titans of industry~ who probably think of themselves as brave ballsy businessmen are actually cowardly pieces of shit when it counts.
September 18, 2025 at 12:26 AM
Shocked but not surprised to learn how many ~titans of industry~ who probably think of themselves as brave ballsy businessmen are actually cowardly pieces of shit when it counts.
Reposted by Erica Kangas
From Wired a few weeks ago. A lot of the people who I was told for years were the adults in the room are being asked how they want to be remembered, and that answer is very clearly "cowards."
www.wired.com/story/uncann...
www.wired.com/story/uncann...
September 18, 2025 at 3:01 AM
From Wired a few weeks ago. A lot of the people who I was told for years were the adults in the room are being asked how they want to be remembered, and that answer is very clearly "cowards."
www.wired.com/story/uncann...
www.wired.com/story/uncann...
Reposted by Erica Kangas
If you’re rich and not a coward, this is what you’d refer to as a “market opportunity” to dominate a media ecosystem that’s about to be covered in grotesque government slop. To be a pop of color in a sea of beige will be easier than ever. People will flock to it. You gotta be a little brave, though.
September 18, 2025 at 12:31 AM
If you’re rich and not a coward, this is what you’d refer to as a “market opportunity” to dominate a media ecosystem that’s about to be covered in grotesque government slop. To be a pop of color in a sea of beige will be easier than ever. People will flock to it. You gotta be a little brave, though.
So interesting to see their AI story unfold 🍿
Some employees who were placed in the talent pool were later offered roles in the company's customer experience team, sources told Business Insider.
Klarna is reassigning engineers and marketers to customer support after its AI bet went too far
Some employees who were placed in the talent pool were later offered roles in the company's customer experience team, sources told Business Insider.
www.businessinsider.com
September 7, 2025 at 5:47 PM
So interesting to see their AI story unfold 🍿
Reposted by Erica Kangas
I talked to Wired for A LONG time about how The Onion works now and also alluded to some very stupid, very fun things coming very soon.
www.wired.com/story/uncann...
www.wired.com/story/uncann...
September 2, 2025 at 4:23 PM
I talked to Wired for A LONG time about how The Onion works now and also alluded to some very stupid, very fun things coming very soon.
www.wired.com/story/uncann...
www.wired.com/story/uncann...
Reposted by Erica Kangas
I just think there's such an illusion of determinism in this industry. As if all things are observed. As if any finding must generalize. As if (and I think this one is a profound one) all software developers are uniform and entirely fungible
"these are the drivers of developer productivity" y'all. Y'ALL. PLEASE.
August 23, 2025 at 10:14 PM
I just think there's such an illusion of determinism in this industry. As if all things are observed. As if any finding must generalize. As if (and I think this one is a profound one) all software developers are uniform and entirely fungible
Reposted by Erica Kangas
New research from Burning Glass concludes that only one in eight nondegree credentials yielded meaningful gains in earnings a year later. Given that returns to short-term certificates typically fade over time, this is not a great outcome. (gift link)
More Workers Are Getting Job-Skill Certificates. They Often Don’t Pay Off.
Many of the thousands of online courses and other credentials employees pursue to boost careers fall short in delivering, a new study finds.
www.wsj.com
August 16, 2025 at 10:20 AM
New research from Burning Glass concludes that only one in eight nondegree credentials yielded meaningful gains in earnings a year later. Given that returns to short-term certificates typically fade over time, this is not a great outcome. (gift link)