Reminder if you use Firefox and haven't done this yet
to disable their A/I chatbot feature (not just removing it from your sidebar):
➡ go to about:config (open a new tab and type it into the address bar)
➡ locate "browser .ml .chat .enabled" (or just search "chat")
➡ toggle to "false"
to disable their A/I chatbot feature (not just removing it from your sidebar):
➡ go to about:config (open a new tab and type it into the address bar)
➡ locate "browser .ml .chat .enabled" (or just search "chat")
➡ toggle to "false"
May 27, 2025 at 2:22 AM
Reminder if you use Firefox and haven't done this yet
to disable their A/I chatbot feature (not just removing it from your sidebar):
➡ go to about:config (open a new tab and type it into the address bar)
➡ locate "browser .ml .chat .enabled" (or just search "chat")
➡ toggle to "false"
to disable their A/I chatbot feature (not just removing it from your sidebar):
➡ go to about:config (open a new tab and type it into the address bar)
➡ locate "browser .ml .chat .enabled" (or just search "chat")
➡ toggle to "false"
A new study resurfaces long-simmering concerns about the lack of transparency in data used to train and validate AI and ML-enabled devices.
FDA authorizes many AI devices for use in kids. Are they validated in a pediatric population?
A new study resurfaces long-simmering concerns about the lack of transparency in data used to train and validate AI and ML-enabled devices.
buff.ly
December 16, 2024 at 9:03 PM
A new study resurfaces long-simmering concerns about the lack of transparency in data used to train and validate AI and ML-enabled devices.
Proper intro!
I'm Dr. Ilenna Jones, postdoctoral Research Fellow at the Harvard Kempner Institute for the Study of Natural and Artificial Intelligence. I ask how #NeuralComputation is enabled by #Dendrite properties using supervised learning tools and methods from #ML and #DeepLearning!
🧠🌻🌳
I'm Dr. Ilenna Jones, postdoctoral Research Fellow at the Harvard Kempner Institute for the Study of Natural and Artificial Intelligence. I ask how #NeuralComputation is enabled by #Dendrite properties using supervised learning tools and methods from #ML and #DeepLearning!
🧠🌻🌳
November 10, 2024 at 9:05 PM
Proper intro!
I'm Dr. Ilenna Jones, postdoctoral Research Fellow at the Harvard Kempner Institute for the Study of Natural and Artificial Intelligence. I ask how #NeuralComputation is enabled by #Dendrite properties using supervised learning tools and methods from #ML and #DeepLearning!
🧠🌻🌳
I'm Dr. Ilenna Jones, postdoctoral Research Fellow at the Harvard Kempner Institute for the Study of Natural and Artificial Intelligence. I ask how #NeuralComputation is enabled by #Dendrite properties using supervised learning tools and methods from #ML and #DeepLearning!
🧠🌻🌳
1.1.1.1 now validates DNSSEC signatures using NIST’s post-quantum ML-DSA-44 algorithm. Here is how we manage 2,420-byte signatures and downgrade risks at scale. https://cfl.re/4yylLLa
1.1.1.1 now supports post-quantum DNSSEC, all 2,420 bytes of it
With ML-DSA-44 validation enabled, 1.1.1.1 lets us test both challenges at Internet scale: carrying larger DNS responses and preventing fallback to conventional signatures.
blog.cloudflare.com
September 10, 2026 at 1:05 PM
1.1.1.1 now validates DNSSEC signatures using NIST’s post-quantum ML-DSA-44 algorithm. Here is how we manage 2,420-byte signatures and downgrade risks at scale. https://cfl.re/4yylLLa
CRISPR, reusable rockets, EUV photolithography, applied cryptography that actually works, photovoltaics at 1/100000th their price of 3 decades ago, scalable high density chemical batteries, just a few (non-ML) things off the top of my head, ironically a few of those enabled the OP’s post lmao
November 28, 2024 at 12:29 PM
CRISPR, reusable rockets, EUV photolithography, applied cryptography that actually works, photovoltaics at 1/100000th their price of 3 decades ago, scalable high density chemical batteries, just a few (non-ML) things off the top of my head, ironically a few of those enabled the OP’s post lmao
Most of my capacity to make pivots in graphics, CV and ML has been enabled by a graduate Numerical Analysis seminar, Hons. Complex Anaylsis Hons. Multivariate Calculus.
Maybe a hot take, but what about the following advice to the next gen:
Don't get an AI degree; the curriculum will be outdated before you graduate. Instead, study math, stats, or physics as your foundation, and stay current with AI through code-focused books, blogs, and papers.
Don't get an AI degree; the curriculum will be outdated before you graduate. Instead, study math, stats, or physics as your foundation, and stay current with AI through code-focused books, blogs, and papers.
February 9, 2025 at 3:42 PM
Most of my capacity to make pivots in graphics, CV and ML has been enabled by a graduate Numerical Analysis seminar, Hons. Complex Anaylsis Hons. Multivariate Calculus.
Our preprint on designing and editing cis-regulatory elements using Ledidi is out! Ledidi turns *any* ML model (or set of models) into a designer of edits to DNA sequences that induce desired characteristics.
Preprint: www.biorxiv.org/content/10.1...
GitHub: github.com/jmschrei/led...
Preprint: www.biorxiv.org/content/10.1...
GitHub: github.com/jmschrei/led...
Programmatic design and editing of cis-regulatory elements
The development of modern genome editing tools has enabled researchers to make such edits with high precision but has left unsolved the problem of designing these edits. As a solution, we propose Ledi...
www.biorxiv.org
April 24, 2025 at 12:59 PM
Our preprint on designing and editing cis-regulatory elements using Ledidi is out! Ledidi turns *any* ML model (or set of models) into a designer of edits to DNA sequences that induce desired characteristics.
Preprint: www.biorxiv.org/content/10.1...
GitHub: github.com/jmschrei/led...
Preprint: www.biorxiv.org/content/10.1...
GitHub: github.com/jmschrei/led...
There were folks on the other site posting at Adobe that they want features like easy ML-enabled layer masking, but those aren't the "AI" features they're getting.
The more I think about it, the more I think it’s going to become increasingly necessary to differentiate from tools that are useful and powered in part by machine learning (content aware fill, advanced CG animation techniques) and generative AI plagiarizing artists at lighting speeds
March 26, 2024 at 10:45 PM
There were folks on the other site posting at Adobe that they want features like easy ML-enabled layer masking, but those aren't the "AI" features they're getting.
PQC in rustls state in a skeet:
ML-KEM is enabled (and prioritized) by default. Initial (and, for now, unstable) support for ML-DSA support is available via the rustls-post-quantum provider (github.com/rustls/rustl...).
ML-KEM is enabled (and prioritized) by default. Initial (and, for now, unstable) support for ML-DSA support is available via the rustls-post-quantum provider (github.com/rustls/rustl...).
post-quantum: add unstable ML-DSA support by djc · Pull Request #2550 · rustls/rustls
Fixes #2417.
github.com
July 22, 2025 at 5:41 PM
PQC in rustls state in a skeet:
ML-KEM is enabled (and prioritized) by default. Initial (and, for now, unstable) support for ML-DSA support is available via the rustls-post-quantum provider (github.com/rustls/rustl...).
ML-KEM is enabled (and prioritized) by default. Initial (and, for now, unstable) support for ML-DSA support is available via the rustls-post-quantum provider (github.com/rustls/rustl...).
Rustls 0.23.44 Released With ML-DSA Certificates Enabled By Default - https://www.phoronix.com/news/Rustls-0.23.44-Released
Rustls 0.23.44 Released With ML-DSA Certificates Enabled By Default
Rustls as the modern TLS library implementation written in the Rust programming language is out with a new feature release. This morning's Rustls 0.23.44 release enables post-quantum secure ML-DSA certificates by default...
www.phoronix.com
September 7, 2026 at 10:18 AM
Rustls 0.23.44 Released With ML-DSA Certificates Enabled By Default - https://www.phoronix.com/news/Rustls-0.23.44-Released
So, coming back to this:
The ToS language people are worrying over a) relates only to the use of specific features within the Zoom application, which are not enabled by default, b) involves opting into ML training, and c) meetings with opt-in have a notification allowing opt-out through leaving.
The ToS language people are worrying over a) relates only to the use of specific features within the Zoom application, which are not enabled by default, b) involves opting into ML training, and c) meetings with opt-in have a notification allowing opt-out through leaving.
Hey uh
I am not a privacy attorney, but I do have a reasonable degree of knowledge about how enterprise software companies do contracts and compliance, and it seems possible that people may be panicking unnecessarily about updates to zoom’s terms of service.
I am not a privacy attorney, but I do have a reasonable degree of knowledge about how enterprise software companies do contracts and compliance, and it seems possible that people may be panicking unnecessarily about updates to zoom’s terms of service.
August 7, 2023 at 6:42 PM
So, coming back to this:
The ToS language people are worrying over a) relates only to the use of specific features within the Zoom application, which are not enabled by default, b) involves opting into ML training, and c) meetings with opt-in have a notification allowing opt-out through leaving.
The ToS language people are worrying over a) relates only to the use of specific features within the Zoom application, which are not enabled by default, b) involves opting into ML training, and c) meetings with opt-in have a notification allowing opt-out through leaving.
I am still pretty convinced that "ML-enabled techno-authoritarianism" is a much bigger threat to <humanity_thrive_percent> than "open-weight models are too good at making bombs." People are already pretty good at making bombs when they're sufficiently motivated.
September 16, 2026 at 2:22 AM
I am still pretty convinced that "ML-enabled techno-authoritarianism" is a much bigger threat to <humanity_thrive_percent> than "open-weight models are too good at making bombs." People are already pretty good at making bombs when they're sufficiently motivated.
My view is that QSAR never delivered consistently and was re-branded as ML (which enabled it to be touted as AI). ML has delivered in areas such as protein fold prediction and image analysis, and I think that the AI label would be more accurate for ML applications like these #chemsky 🧪
November 23, 2024 at 8:39 AM
My view is that QSAR never delivered consistently and was re-branded as ML (which enabled it to be touted as AI). ML has delivered in areas such as protein fold prediction and image analysis, and I think that the AI label would be more accurate for ML applications like these #chemsky 🧪
hey other firefox users, the AI chatbot doesn't seem completely turned off if you just turn it off in the sidebar. if you look up ml (machine learning) in a new tab called about:config, you'll likely see a lot of different things still on
August 15, 2025 at 2:53 PM
hey other firefox users, the AI chatbot doesn't seem completely turned off if you just turn it off in the sidebar. if you look up ml (machine learning) in a new tab called about:config, you'll likely see a lot of different things still on
Structure-enabled enzyme function prediction unveils elusive terpenoid biosynthesis in archaea [updated]
Archaea terpene synthases ID'd via ML.
Archaea terpene synthases ID'd via ML.
April 30, 2025 at 7:00 AM
Structure-enabled enzyme function prediction unveils elusive terpenoid biosynthesis in archaea [updated]
Archaea terpene synthases ID'd via ML.
Archaea terpene synthases ID'd via ML.
Not surprising, but certainly a blaring alarm for technology transfer. AI-enabled applications will need to show obvious value and respect user concerns (e.g. privacy) #AI #ML
April 7, 2025 at 9:40 PM
The FDA's most recent guidance document on Predetermined Change Control Plans for AI-enabled devices does an admirable job of recognizing the complexity inherent in implementing AI in healthcare settings, allowing us to practice the MLOps essentials needed for safe AI.
tinyurl.com/fda-pccp-gui...
tinyurl.com/fda-pccp-gui...
Predetermined Change Control Plan for Artificial Intelligence-Enabled
Marketing Submission Recommendations for Predetermined Change Control Plans for AI/ML-Enabled Device Software Functions
www.fda.gov
December 4, 2024 at 12:34 AM
The FDA's most recent guidance document on Predetermined Change Control Plans for AI-enabled devices does an admirable job of recognizing the complexity inherent in implementing AI in healthcare settings, allowing us to practice the MLOps essentials needed for safe AI.
tinyurl.com/fda-pccp-gui...
tinyurl.com/fda-pccp-gui...
Would you also, @webdevs.firefox.com, care to go to about:config, search for ".ml.", look at whether all of those that are
1. Named something along the lines of "enabled", and
2. Are set to "true"
3. Are not presented as diverting from the default setting
…and give your personal assessment… 1/2
1. Named something along the lines of "enabled", and
2. Are set to "true"
3. Are not presented as diverting from the default setting
…and give your personal assessment… 1/2
December 19, 2025 at 11:30 PM
Would you also, @webdevs.firefox.com, care to go to about:config, search for ".ml.", look at whether all of those that are
1. Named something along the lines of "enabled", and
2. Are set to "true"
3. Are not presented as diverting from the default setting
…and give your personal assessment… 1/2
1. Named something along the lines of "enabled", and
2. Are set to "true"
3. Are not presented as diverting from the default setting
…and give your personal assessment… 1/2
Just trying to promote my talk in Chile 😅 If you're interested in fraud prevention or ML-enabled software architecture, it might be worth a look.
To support, you can like this LI post -> www.linkedin.com/posts/jose-b...
Thanks a lot for your attention! 🙇♀️
To support, you can like this LI post -> www.linkedin.com/posts/jose-b...
Thanks a lot for your attention! 🙇♀️
March 17, 2026 at 6:19 PM
Just trying to promote my talk in Chile 😅 If you're interested in fraud prevention or ML-enabled software architecture, it might be worth a look.
To support, you can like this LI post -> www.linkedin.com/posts/jose-b...
Thanks a lot for your attention! 🙇♀️
To support, you can like this LI post -> www.linkedin.com/posts/jose-b...
Thanks a lot for your attention! 🙇♀️
Firefox has enabled AI, they call it ML = Machine Learning. To turn it off, go to Advanced Preferences which is the old "about:config" page. Type that into the address bar. Confirm that you want to see the scary settings. Type "\.ml\.*enable" in the search bar, and set to false where appropriate.
August 27, 2025 at 6:34 PM
Firefox has enabled AI, they call it ML = Machine Learning. To turn it off, go to Advanced Preferences which is the old "about:config" page. Type that into the address bar. Confirm that you want to see the scary settings. Type "\.ml\.*enable" in the search bar, and set to false where appropriate.
I found a property when searching browser .ml that might be worth turning off as well: "browser . ml . enabled". I'm disabling it, but do you think this is needed especially since the one you gave was specifically for the chat feature?
May 27, 2025 at 7:44 PM
I found a property when searching browser .ml that might be worth turning off as well: "browser . ml . enabled". I'm disabling it, but do you think this is needed especially since the one you gave was specifically for the chat feature?
@lukasmsigmund.bsky.social has now enabled us not only to predict reaction selectivity with ML methods, but also the reaction feasibility, very nice work! Thanks for another great collaboration @valencekjell.com doi.org/10.26434/che...
Predicting Reaction Feasibility and Selectivity of Aromatic C–H Thianthrenation with a QM-ML Hybrid Approach
The direct thianthrenation of aromatic C–H bonds is a valuable late-stage functionalization strategy that can assist, for example, the development of new drugs. We herein present a predictive computat...
doi.org
August 22, 2025 at 9:16 AM
@lukasmsigmund.bsky.social has now enabled us not only to predict reaction selectivity with ML methods, but also the reaction feasibility, very nice work! Thanks for another great collaboration @valencekjell.com doi.org/10.26434/che...
Though even if it doesn’t for mine, it may be enabled on other’s Apple Watches which isn’t comfortable either.
In general I feel Apple is pushing “Apple Intelligence“ far too much. Like there is upsides to some aspects of ML, especially in a11y, but this ain’t it.
In general I feel Apple is pushing “Apple Intelligence“ far too much. Like there is upsides to some aspects of ML, especially in a11y, but this ain’t it.
September 17, 2026 at 12:33 PM
Though even if it doesn’t for mine, it may be enabled on other’s Apple Watches which isn’t comfortable either.
In general I feel Apple is pushing “Apple Intelligence“ far too much. Like there is upsides to some aspects of ML, especially in a11y, but this ain’t it.
In general I feel Apple is pushing “Apple Intelligence“ far too much. Like there is upsides to some aspects of ML, especially in a11y, but this ain’t it.
i'm pretty sure I'd classify that under "ML-enabled techno-authoritarianism" but I might be missing some /s tags here lol
September 16, 2026 at 2:54 AM
i'm pretty sure I'd classify that under "ML-enabled techno-authoritarianism" but I might be missing some /s tags here lol
We should also remember the many local activists who pushed back against the democratizing movements of their day, the "constant gardeners" of historical bigotries like Walter Huss, the chair of the Oregon Republican Party in 1978-9. Stories like that matter. sethcotlar.substack.com/p/on-this-ml...
On this MLK Day, we should remember the people who in 1989 wanted to take MLK's name off a Portland street, but for totally not-racist reasons!
On Walter Huss's lifelong hatred of MLK, the many "totally not racist" Oregonians who enabled it, and the small group of antifascist activists and researchers who told the truth about it
sethcotlar.substack.com
June 25, 2024 at 2:39 PM
We should also remember the many local activists who pushed back against the democratizing movements of their day, the "constant gardeners" of historical bigotries like Walter Huss, the chair of the Oregon Republican Party in 1978-9. Stories like that matter. sethcotlar.substack.com/p/on-this-ml...