Happy to see this short piece come out, titled "The Ouroboros of Synthetic Data: On Undead Labor and Capitalist Surrealism," via In Progress.
www.globalemergentmedia.com/in-progress/...
www.globalemergentmedia.com/in-progress/...
The Ouroboros of Synthetic Data: On Undead Labor and Capitalist Surrealism
members
www.globalemergentmedia.com
November 3, 2026 at 4:12 PM
Happy to see this short piece come out, titled "The Ouroboros of Synthetic Data: On Undead Labor and Capitalist Surrealism," via In Progress.
www.globalemergentmedia.com/in-progress/...
www.globalemergentmedia.com/in-progress/...
I have tried to tackle this head on pretty much every reply and also in the past: the tool itself is just a tool, one that can (and hopefully will) work without stealing data or oppressing anyone by default (synthetic data is a thing that does work), I do not see any technology as just “evil” tbh
November 3, 2026 at 10:19 PM
I have tried to tackle this head on pretty much every reply and also in the past: the tool itself is just a tool, one that can (and hopefully will) work without stealing data or oppressing anyone by default (synthetic data is a thing that does work), I do not see any technology as just “evil” tbh
Calibre MCP is live. Connect Claude Code, Cursor, Codex or Copilot to your Calibre data and ask about any page: “Why is LCP slow on /checkout?” or “Did yesterday’s deploy make anything slower?”
Your agent finds the cause in your RUM, CrUX and Synthetic data, then fixes it in your code.
Your agent finds the cause in your RUM, CrUX and Synthetic data, then fixes it in your code.
Calibre MCP: connect your coding agent to your performance data
Connect Claude Code, Cursor, Codex or GitHub Copilot to Calibre and ask about your RUM, CrUX and Synthetic data in plain language. Read-only and included on every plan.
calibreapp.com
November 3, 2026 at 11:52 PM
Calibre MCP is live. Connect Claude Code, Cursor, Codex or Copilot to your Calibre data and ask about any page: “Why is LCP slow on /checkout?” or “Did yesterday’s deploy make anything slower?”
Your agent finds the cause in your RUM, CrUX and Synthetic data, then fixes it in your code.
Your agent finds the cause in your RUM, CrUX and Synthetic data, then fixes it in your code.
So they no longer need internet data and can use synthetic data is that what this is ?
November 2, 2026 at 5:14 PM
So they no longer need internet data and can use synthetic data is that what this is ?
Demand for defunct companies’ data emerged suddenly last year, as AI training labs hit the limits of public and synthetic datasets.
AI labs are buying data from dead startups to train their models
Most startups fail. A new industry is giving their data an afterlife in AI.
trib.al
November 4, 2026 at 10:42 AM
Demand for defunct companies’ data emerged suddenly last year, as AI training labs hit the limits of public and synthetic datasets.
…independent auditing of data use or of data hygiene. This is all purely based on trust and reputation damage if caught. All data is retained for between 30 days and 5 years, regardless of contract. And there are other uses of data, such as to generate synthetic training data, that might be…
November 4, 2026 at 9:09 PM
…independent auditing of data use or of data hygiene. This is all purely based on trust and reputation damage if caught. All data is retained for between 30 days and 5 years, regardless of contract. And there are other uses of data, such as to generate synthetic training data, that might be…
Happy to share a new preprint led by Lena Wang exploring the use of synthetic MPRA data to improve the performance of sequence to function models, achieving better generalization to unseen genes:
www.biorxiv.org/content/10.6...
www.biorxiv.org/content/10.6...
Synthetic MPRA perturbation supervision improves variant-effect prediction in sequence-to-function models
Deep learning sequence-to-function (S2F) models learn to predict functional readouts, such as gene expression, from genetic sequence. However, these models are trained predominantly on the reference g...
www.biorxiv.org
November 4, 2026 at 2:41 PM
Happy to share a new preprint led by Lena Wang exploring the use of synthetic MPRA data to improve the performance of sequence to function models, achieving better generalization to unseen genes:
www.biorxiv.org/content/10.6...
www.biorxiv.org/content/10.6...
Synthetic data in ai: an essential guide for 2026
https://beyondaibooks.blogspot.com/2026/11/synthetic-data-in-ai-essential-guide.html
https://beyondaibooks.blogspot.com/2026/11/synthetic-data-in-ai-essential-guide.html
November 5, 2026 at 6:30 AM
Synthetic data in ai: an essential guide for 2026
https://beyondaibooks.blogspot.com/2026/11/synthetic-data-in-ai-essential-guide.html
https://beyondaibooks.blogspot.com/2026/11/synthetic-data-in-ai-essential-guide.html
Synthetic Audiences, built from real data. Say hi to the newest feature in YouScan 🚀
Turn real social conversations into personas and see how they're likely to react to your next post before you spend a cent on it.
Try it out → youscan.io/synthetic-au...
Turn real social conversations into personas and see how they're likely to react to your next post before you spend a cent on it.
Try it out → youscan.io/synthetic-au...
Synthetic Audiences: Personas Built from Real Data
Test messages, creatives, and concepts with Synthetic Audiences based on real social mentions. See how your audience is likely to react to them before launch.
youscan.io
November 3, 2026 at 1:19 PM
Synthetic Audiences, built from real data. Say hi to the newest feature in YouScan 🚀
Turn real social conversations into personas and see how they're likely to react to your next post before you spend a cent on it.
Try it out → youscan.io/synthetic-au...
Turn real social conversations into personas and see how they're likely to react to your next post before you spend a cent on it.
Try it out → youscan.io/synthetic-au...
November 4, 2026 at 11:23 PM
It’s a lot about the data. How can you keep a “simple” classifier neural network up to date in a post gpt world (as a language continuously changes)? With synthetic training and evaluation data for example, but that’s already a bias.
November 6, 2026 at 2:21 PM
It’s a lot about the data. How can you keep a “simple” classifier neural network up to date in a post gpt world (as a language continuously changes)? With synthetic training and evaluation data for example, but that’s already a bias.
🫁 November is #LungCancerAwarenessMonth!
#SYNTHIA explores how #synthetic imaging and clinical data can improve #lungcancer detection, treatment monitoring and outcome prediction while protecting patient privacy.
Learn more: www.ihi-synthia.eu/use-cases/lu...
#IHITransformingHealth #HorizonEU
#SYNTHIA explores how #synthetic imaging and clinical data can improve #lungcancer detection, treatment monitoring and outcome prediction while protecting patient privacy.
Learn more: www.ihi-synthia.eu/use-cases/lu...
#IHITransformingHealth #HorizonEU
November 4, 2026 at 10:35 AM
🫁 November is #LungCancerAwarenessMonth!
#SYNTHIA explores how #synthetic imaging and clinical data can improve #lungcancer detection, treatment monitoring and outcome prediction while protecting patient privacy.
Learn more: www.ihi-synthia.eu/use-cases/lu...
#IHITransformingHealth #HorizonEU
#SYNTHIA explores how #synthetic imaging and clinical data can improve #lungcancer detection, treatment monitoring and outcome prediction while protecting patient privacy.
Learn more: www.ihi-synthia.eu/use-cases/lu...
#IHITransformingHealth #HorizonEU
🆕 Calibre MCP: ask your coding agent "Why is LCP slow on /checkout?" and it finds the answer in your RUM, CrUX and Synthetic data, then fixes it in your code.
Read-only, on every plan.
calibreapp.com/changelog/ar...
Read-only, on every plan.
calibreapp.com/changelog/ar...
Calibre MCP: connect your coding agent to your performance data | Calibre
Connect Claude Code, Cursor, Codex or GitHub Copilot to Calibre and ask about your RUM, CrUX and Synthetic data in plain language. Read-only and included on every plan.
calibreapp.com
November 3, 2026 at 6:48 AM
🆕 Calibre MCP: ask your coding agent "Why is LCP slow on /checkout?" and it finds the answer in your RUM, CrUX and Synthetic data, then fixes it in your code.
Read-only, on every plan.
calibreapp.com/changelog/ar...
Read-only, on every plan.
calibreapp.com/changelog/ar...
My amazing colleagues are hosting a meet up in the Meta offices on NYC to talk about synthetic data and reliable AI on Nov 16, check it out! And sign up if you're interested in attending.
luma.com/gfkzrpx3
luma.com/gfkzrpx3
Synthetic data for reliable AI · Luma
Synthetic data can unlocks unlimited targeted data for training and evaluating AI systems. From distillation to synthetic environments, synthetic data can be…
luma.com
October 30, 2026 at 9:32 PM
My amazing colleagues are hosting a meet up in the Meta offices on NYC to talk about synthetic data and reliable AI on Nov 16, check it out! And sign up if you're interested in attending.
luma.com/gfkzrpx3
luma.com/gfkzrpx3
🎤 Vagrant Gautam @dippedrusk.com
“Measurement and impact in bias research”
After nearly a decade of bias research, do we agree on how to measure it? Vagrant revisits debates such as intrinsic vs. extrinsic bias, natural vs. synthetic data, and where the field should go next.
“Measurement and impact in bias research”
After nearly a decade of bias research, do we agree on how to measure it? Vagrant revisits debates such as intrinsic vs. extrinsic bias, natural vs. synthetic data, and where the field should go next.
November 2, 2026 at 12:34 PM
🎤 Vagrant Gautam @dippedrusk.com
“Measurement and impact in bias research”
After nearly a decade of bias research, do we agree on how to measure it? Vagrant revisits debates such as intrinsic vs. extrinsic bias, natural vs. synthetic data, and where the field should go next.
“Measurement and impact in bias research”
After nearly a decade of bias research, do we agree on how to measure it? Vagrant revisits debates such as intrinsic vs. extrinsic bias, natural vs. synthetic data, and where the field should go next.
Super cool demo she just did showed how 3D data can cleanly and quickly be edited in VR.
Data doesn't have to be from a telescope - this can also handle synthetic data from computer models.
Data doesn't have to be from a telescope - this can also handle synthetic data from computer models.
November 3, 2026 at 12:38 AM
Super cool demo she just did showed how 3D data can cleanly and quickly be edited in VR.
Data doesn't have to be from a telescope - this can also handle synthetic data from computer models.
Data doesn't have to be from a telescope - this can also handle synthetic data from computer models.
Data could be used in the post-training stage. To provide samples for fine-tuning. In a RAG. To generate synthetic data. And who knows how else. All of this sounds compatible with a “no training” agreement. You take your chances. Like with Meta.
November 4, 2026 at 9:09 PM
Data could be used in the post-training stage. To provide samples for fine-tuning. In a RAG. To generate synthetic data. And who knows how else. All of this sounds compatible with a “no training” agreement. You take your chances. Like with Meta.
As far as I can tell from all the reports, the posts didn't go through the data centers - the programs running there generated fake likes and reposts.
A lot of that synthetic feedback boosted Elon Musk, Tucker Carlson, and other right-wing nutjobs.
A lot of that synthetic feedback boosted Elon Musk, Tucker Carlson, and other right-wing nutjobs.
November 4, 2026 at 5:53 PM
As far as I can tell from all the reports, the posts didn't go through the data centers - the programs running there generated fake likes and reposts.
A lot of that synthetic feedback boosted Elon Musk, Tucker Carlson, and other right-wing nutjobs.
A lot of that synthetic feedback boosted Elon Musk, Tucker Carlson, and other right-wing nutjobs.
📢VoxLution: Real-world scans are noisy & partial, so 3D scene generators often rely on synthetic data.
Daoyi Gao & Xinyi Zhang introduce a sparse voxel cascaded transformer trained on real scans, generating high-fidelity realistic scenes!
Check it out: daoyig.github.io/VoxLution/
Daoyi Gao & Xinyi Zhang introduce a sparse voxel cascaded transformer trained on real scans, generating high-fidelity realistic scenes!
Check it out: daoyig.github.io/VoxLution/
November 2, 2026 at 5:27 PM
📢VoxLution: Real-world scans are noisy & partial, so 3D scene generators often rely on synthetic data.
Daoyi Gao & Xinyi Zhang introduce a sparse voxel cascaded transformer trained on real scans, generating high-fidelity realistic scenes!
Check it out: daoyig.github.io/VoxLution/
Daoyi Gao & Xinyi Zhang introduce a sparse voxel cascaded transformer trained on real scans, generating high-fidelity realistic scenes!
Check it out: daoyig.github.io/VoxLution/
Synthetic Data Generation Market Size, Forecast Analysis - 2035
✔For Additional Details, Visit Here: www.marketresearchfuture.com/reports/synt...
✔For Additional Details, Visit Here: www.marketresearchfuture.com/reports/synt...
Synthetic Data Generation Market Size, Forecast Analysis - 2035
Synthetic Data Generation Market is Estimated to Reach USD 34.62 Billion By 2035, Growing at a CAGR of 46.30% During 2025 - 2035 | Demand For Diverse Datasets
www.marketresearchfuture.com
November 4, 2026 at 5:30 AM
Synthetic Data Generation Market Size, Forecast Analysis - 2035
✔For Additional Details, Visit Here: www.marketresearchfuture.com/reports/synt...
✔For Additional Details, Visit Here: www.marketresearchfuture.com/reports/synt...
DeepSeek V3 & R1 Off-Peak UTC Window Calculator (50%–75% Batch API Savings)
Schedule non-urgent RAG indexing, batch classification, and synthetic data jobs during 16:30–00:30 UTC (12:30 PM–8:30 PM EST) to capture DeepSeek off-peak pricing discounts. 🧵👇
#AI #LLM #DevTools #SaaS #BuildInPublic
Schedule non-urgent RAG indexing, batch classification, and synthetic data jobs during 16:30–00:30 UTC (12:30 PM–8:30 PM EST) to capture DeepSeek off-peak pricing discounts. 🧵👇
#AI #LLM #DevTools #SaaS #BuildInPublic
November 4, 2026 at 5:54 PM
DeepSeek V3 & R1 Off-Peak UTC Window Calculator (50%–75% Batch API Savings)
Schedule non-urgent RAG indexing, batch classification, and synthetic data jobs during 16:30–00:30 UTC (12:30 PM–8:30 PM EST) to capture DeepSeek off-peak pricing discounts. 🧵👇
#AI #LLM #DevTools #SaaS #BuildInPublic
Schedule non-urgent RAG indexing, batch classification, and synthetic data jobs during 16:30–00:30 UTC (12:30 PM–8:30 PM EST) to capture DeepSeek off-peak pricing discounts. 🧵👇
#AI #LLM #DevTools #SaaS #BuildInPublic
Just plug it straight into my veins.
November 3, 2026 at 1:30 PM
Just plug it straight into my veins.
This.
Anti AI don't realise.
We're their greatest friends.
There's nothing more we'd want than OpenAI/Anthropic to make their models opensource, and have used either open data/synthetic data (who wants reddit mods in the data?!) or paid for the art/data!
But this is the lemons we have...
Anti AI don't realise.
We're their greatest friends.
There's nothing more we'd want than OpenAI/Anthropic to make their models opensource, and have used either open data/synthetic data (who wants reddit mods in the data?!) or paid for the art/data!
But this is the lemons we have...
I think Sam may be an instrument of divine providence to give both sides a common enemy to unify around
October 31, 2026 at 4:09 AM
This.
Anti AI don't realise.
We're their greatest friends.
There's nothing more we'd want than OpenAI/Anthropic to make their models opensource, and have used either open data/synthetic data (who wants reddit mods in the data?!) or paid for the art/data!
But this is the lemons we have...
Anti AI don't realise.
We're their greatest friends.
There's nothing more we'd want than OpenAI/Anthropic to make their models opensource, and have used either open data/synthetic data (who wants reddit mods in the data?!) or paid for the art/data!
But this is the lemons we have...
17/18 We have moved beyond skepticism toward synthetic data, such as the +200M structures predicted by AlphaFold, to directly incorporating these predictions as a major foundation for our algorithms.
In 2022, we remained cautious but optimistic about ...
In 2022, we remained cautious but optimistic about ...
November 4, 2026 at 10:02 PM
17/18 We have moved beyond skepticism toward synthetic data, such as the +200M structures predicted by AlphaFold, to directly incorporating these predictions as a major foundation for our algorithms.
In 2022, we remained cautious but optimistic about ...
In 2022, we remained cautious but optimistic about ...