This article by @deenamousa.com about the problem with radiology screening shows this very well:
www.understandingai.org/p/ai-isnt-re...
Besides, you can't teach someone English using only appliance repair manuals and then expect them to write a decent sonnet.
www.understandingai.org/p/ai-isnt-re...
Besides, you can't teach someone English using only appliance repair manuals and then expect them to write a decent sonnet.
November 12, 2025 at 1:57 PM
This article by @deenamousa.com about the problem with radiology screening shows this very well:
www.understandingai.org/p/ai-isnt-re...
Besides, you can't teach someone English using only appliance repair manuals and then expect them to write a decent sonnet.
www.understandingai.org/p/ai-isnt-re...
Besides, you can't teach someone English using only appliance repair manuals and then expect them to write a decent sonnet.
i need a lot more evidence. i need benchmarks on games that are by devs i trust. i need to talk to the CTO and understand how it is that their product automatically ports things so that they're faster than running native. don't worry -- i won't embarrass myself. i've done this at google and unity.
November 12, 2025 at 1:45 PM
i need a lot more evidence. i need benchmarks on games that are by devs i trust. i need to talk to the CTO and understand how it is that their product automatically ports things so that they're faster than running native. don't worry -- i won't embarrass myself. i've done this at google and unity.
También encontré un canal decente de reseñas de tablets (se llama My Next Tablet), que costó mucho más de lo que pensaría. Y tampoco es que sea súper técnico, pero al menos ponen un par de benchmarks comparando dispositivos similares y no se pasan media reseña hablando de los altavoces o la cámara.
November 12, 2025 at 1:43 PM
También encontré un canal decente de reseñas de tablets (se llama My Next Tablet), que costó mucho más de lo que pensaría. Y tampoco es que sea súper técnico, pero al menos ponen un par de benchmarks comparando dispositivos similares y no se pasan media reseña hablando de los altavoces o la cámara.
In Texas, standardized testing is not only not developmentally appropriate, but also gets rewritten/reconfigured every few years, as soon as students begin to make gains, making it clear the whole point is to destabilize public education, not get benchmarks to understand schools' effectiveness.
November 12, 2025 at 1:33 PM
In Texas, standardized testing is not only not developmentally appropriate, but also gets rewritten/reconfigured every few years, as soon as students begin to make gains, making it clear the whole point is to destabilize public education, not get benchmarks to understand schools' effectiveness.
📊 Consider structured exposure (call spreads or collars) over outright OTM calls given rich speculation, and monitor front-end IV into any rumored announcement windows. For pairs, maintain flexibility: if AMD’s MI450 claims gain third-party benchmarks, AMD/NVDA relative value could pivot (9/10)
November 12, 2025 at 1:24 PM
📊 Consider structured exposure (call spreads or collars) over outright OTM calls given rich speculation, and monitor front-end IV into any rumored announcement windows. For pairs, maintain flexibility: if AMD’s MI450 claims gain third-party benchmarks, AMD/NVDA relative value could pivot (9/10)
Canadian Family Offices: Subscriber-Only 2025 Multi-Family Office Study Offers Fresh Benchmarks For Growth, Fees, And Governance (Nov. 12, 2025)
The recent insights on the 2025 Multi-Family Office (MFO) Study emphasize the evolving landscape of MFO services as firms navigate new market dynamics.…
The recent insights on the 2025 Multi-Family Office (MFO) Study emphasize the evolving landscape of MFO services as firms navigate new market dynamics.…
Canadian Family Offices: Subscriber-Only 2025 Multi-Family Office Study Offers Fresh Benchmarks For Growth, Fees, And Governance (Nov. 12, 2025)
The recent insights on the 2025 Multi-Family Office (MFO) Study emphasize the evolving landscape of MFO services as firms navigate new market dynamics. With 2025 planning cycles approaching, the study presents crucial benchmarks on client growth, service offerings, and staffing models amid fluctuating market conditions and heightened client expectations. Critical areas of focus include fee structures, governance frameworks, and integrated solutions for tax, estate, and risk management, all while leveraging technology for improved efficiency. By examining staffing dynamics and client complexity, firms can refine their service models to ensure profitability and meet increasing standards for fiduciary oversight. Actionable strategies for leaders include rigorous fee benchmarking, deliberate service alignment, and strengthening governance, ultimately enhancing client trust and operational effectiveness in the MFO sector.
wealthstrategiesjournal.com
November 12, 2025 at 1:08 PM
Canadian Family Offices: Subscriber-Only 2025 Multi-Family Office Study Offers Fresh Benchmarks For Growth, Fees, And Governance (Nov. 12, 2025)
The recent insights on the 2025 Multi-Family Office (MFO) Study emphasize the evolving landscape of MFO services as firms navigate new market dynamics.…
The recent insights on the 2025 Multi-Family Office (MFO) Study emphasize the evolving landscape of MFO services as firms navigate new market dynamics.…
some context: 80 layers is very deep for a small model
what if you took Qwen3-4B with 36 layers and looped it 4x? That's somewhat analogous to 144 layers
it won't do better on knowledge benchmarks, but it certainly gets us closer to that coveted cognitive core that goes external for knowledge
what if you took Qwen3-4B with 36 layers and looped it 4x? That's somewhat analogous to 144 layers
it won't do better on knowledge benchmarks, but it certainly gets us closer to that coveted cognitive core that goes external for knowledge
November 12, 2025 at 12:49 PM
some context: 80 layers is very deep for a small model
what if you took Qwen3-4B with 36 layers and looped it 4x? That's somewhat analogous to 144 layers
it won't do better on knowledge benchmarks, but it certainly gets us closer to that coveted cognitive core that goes external for knowledge
what if you took Qwen3-4B with 36 layers and looped it 4x? That's somewhat analogous to 144 layers
it won't do better on knowledge benchmarks, but it certainly gets us closer to that coveted cognitive core that goes external for knowledge
At DevCon in Nov 18-19, Ray Myers will go over emerging strategies for improving the accuracy of coding agents on real codebases, benchmarks such as SWE-bench that evaluate our progress, and their limitations.
🧵 2/3
🧵 2/3
November 12, 2025 at 12:30 PM
At DevCon in Nov 18-19, Ray Myers will go over emerging strategies for improving the accuracy of coding agents on real codebases, benchmarks such as SWE-bench that evaluate our progress, and their limitations.
🧵 2/3
🧵 2/3
Les smartphones Android trichent tellement dans les benchmarks que les développeurs en profitent pour accélérer les émulateurs http:// dlvr.it/TPD6zk # smartphones # Android
Interest | Match | Feed
Interest | Match | Feed
Origin
social.macg.co
November 12, 2025 at 12:20 PM
Les smartphones Android trichent tellement dans les benchmarks que les développeurs en profitent pour accélérer les émulateurs http:// dlvr.it/TPD6zk # smartphones # Android
Interest | Match | Feed
Interest | Match | Feed
Origin
social.macg.co
November 12, 2025 at 12:07 PM
Jegliche KI Benchmarks sind unbrauchbar.
November 12, 2025 at 11:59 AM
Jegliche KI Benchmarks sind unbrauchbar.
Yeah, I was going to comment similar. Most of the heaviest hitters for the GBA especially are just out of reach, and increasingly so for the DS as that gets more notice and nostalgia. So I'm very much for something like this, just best to wait and see for reviews and benchmarks before leaping on.
November 12, 2025 at 11:47 AM
Yeah, I was going to comment similar. Most of the heaviest hitters for the GBA especially are just out of reach, and increasingly so for the DS as that gets more notice and nostalgia. So I'm very much for something like this, just best to wait and see for reviews and benchmarks before leaping on.
Post an album cover once a day for a year. Only albums that you currently (or at some point in your life) kept on repeat.
Life is a journey!
Share some of the benchmarks along your path!
#PaulSimon
#Music #MusicSky #Album #MusicChallenge
[82-365]
Life is a journey!
Share some of the benchmarks along your path!
#PaulSimon
#Music #MusicSky #Album #MusicChallenge
[82-365]
November 12, 2025 at 11:07 AM
Post an album cover once a day for a year. Only albums that you currently (or at some point in your life) kept on repeat.
Life is a journey!
Share some of the benchmarks along your path!
#PaulSimon
#Music #MusicSky #Album #MusicChallenge
[82-365]
Life is a journey!
Share some of the benchmarks along your path!
#PaulSimon
#Music #MusicSky #Album #MusicChallenge
[82-365]
Influencer Marketing Benchmarks: A Pricing Guide from Wix
quasa.io/media/influe...
quasa.io/media/influe...
Influencer Marketing Benchmarks: A Pricing Guide from Wix
Sarah Adam, the Head of Influencer Marketing at Wix, has published an incredibly valuable pricing guide based on her extensive experience working with influencers
quasa.io
November 12, 2025 at 10:57 AM
Influencer Marketing Benchmarks: A Pricing Guide from Wix
quasa.io/media/influe...
quasa.io/media/influe...
Have you had an uneasy feeling about the clear gap between how LLMs perform on benchmarks and their real-life capabilities?
If you're curious about the messy truth behind current LLM assessment, catch my @ndcconferences.com AI keynote tomorrow at Rebel, Oslo.
ndc-ai.com/agenda/can-y...
If you're curious about the messy truth behind current LLM assessment, catch my @ndcconferences.com AI keynote tomorrow at Rebel, Oslo.
ndc-ai.com/agenda/can-y...
Keynote: Can you trust your (large language) model? | NDC AI 2025
Machine learning algorithms are marvellous things: models that can do a bunch of tedious and complex tasks for us, all with a high degree of accuracy. But how do we really know whether the outputs of ...
ndc-ai.com
November 12, 2025 at 10:54 AM
Have you had an uneasy feeling about the clear gap between how LLMs perform on benchmarks and their real-life capabilities?
If you're curious about the messy truth behind current LLM assessment, catch my @ndcconferences.com AI keynote tomorrow at Rebel, Oslo.
ndc-ai.com/agenda/can-y...
If you're curious about the messy truth behind current LLM assessment, catch my @ndcconferences.com AI keynote tomorrow at Rebel, Oslo.
ndc-ai.com/agenda/can-y...
"Europe is preparing to roll back parts of its landmark digital rules, long seen as global benchmarks for privacy and AI" @ramshajahangir.bsky.social writes, as EU Commission prepares to unveil the “Digital Omnibus” which could reshape GDPR, AI Act, and more. 1/4 www.techpolicy.press/eu-set-the-g...
EU Set the Global Standard on Privacy and AI. Now It’s Pulling Back | TechPolicy.Press
The draft Digital Omnibus could weaken core data protections and give tech companies more leeway in using European data, reports Ramsha Jahangir.
www.techpolicy.press
November 12, 2025 at 10:26 AM
"Europe is preparing to roll back parts of its landmark digital rules, long seen as global benchmarks for privacy and AI" @ramshajahangir.bsky.social writes, as EU Commission prepares to unveil the “Digital Omnibus” which could reshape GDPR, AI Act, and more. 1/4 www.techpolicy.press/eu-set-the-g...
Build conversation depth: reply to the replies and ask follow‑ups like “what’s the shortest project you ever hired for?” and “how did you measure its success?”
What’s one micro‑job outcome you’d hire or accept this year, and why? Let’s share ideas and benchmarks.
What’s one micro‑job outcome you’d hire or accept this year, and why? Let’s share ideas and benchmarks.
November 12, 2025 at 10:14 AM
Build conversation depth: reply to the replies and ask follow‑ups like “what’s the shortest project you ever hired for?” and “how did you measure its success?”
What’s one micro‑job outcome you’d hire or accept this year, and why? Let’s share ideas and benchmarks.
What’s one micro‑job outcome you’d hire or accept this year, and why? Let’s share ideas and benchmarks.
Must reads for #genai people who look at #llm #benchmarks
https://arxiv.org/pdf/2511.04703
https://open.substack.com/pub/oneusefulthing/p/giving-your-ai-a-job-interview
https://arxiv.org/pdf/2511.04703
https://open.substack.com/pub/oneusefulthing/p/giving-your-ai-a-job-interview
Giving your AI a Job Interview
As AI advice becomes more important, we are going to need to get better at assessing it
www.oneusefulthing.org
November 12, 2025 at 8:34 AM
Must reads for #genai people who look at #llm #benchmarks
https://arxiv.org/pdf/2511.04703
https://open.substack.com/pub/oneusefulthing/p/giving-your-ai-a-job-interview
https://arxiv.org/pdf/2511.04703
https://open.substack.com/pub/oneusefulthing/p/giving-your-ai-a-job-interview
Steve's journey from gaming benchmarks to investigative journalist looking into tech corp corruption should resonate with us all tbh
The next round of special reports delve into US export controls, what we think is corruption between governments and tech companies, reckless data center expansion, and more. Here's the plan: www.youtube.com/watch?v=qG4e...
Contacted by the US Secret Service & the AI Surveillance Center Dystopia
YouTube video by Gamers Nexus
www.youtube.com
November 12, 2025 at 8:01 AM
Steve's journey from gaming benchmarks to investigative journalist looking into tech corp corruption should resonate with us all tbh
#Music #MusicSky #Album
#MusicChallenge
Post an album cover once a day for a year. Only albums that you currently ( or at some point in your life) kept on repeat.
Life is a journey!
Share some of the benchmarks along your path!
#PrettyMaids
Future World
49/365
#MusicChallenge
Post an album cover once a day for a year. Only albums that you currently ( or at some point in your life) kept on repeat.
Life is a journey!
Share some of the benchmarks along your path!
#PrettyMaids
Future World
49/365
November 12, 2025 at 4:54 AM
#Music #MusicSky #Album
#MusicChallenge
Post an album cover once a day for a year. Only albums that you currently ( or at some point in your life) kept on repeat.
Life is a journey!
Share some of the benchmarks along your path!
#PrettyMaids
Future World
49/365
#MusicChallenge
Post an album cover once a day for a year. Only albums that you currently ( or at some point in your life) kept on repeat.
Life is a journey!
Share some of the benchmarks along your path!
#PrettyMaids
Future World
49/365
One of the most reliable benchmarks I have ever encountered for judging the character of someone has been: how they treat their furry frens. People can be consistent fuck ups in just about every other way, but if they take good care of their fren, I'm gonna think "jury still out on that person".
November 12, 2025 at 4:25 AM
One of the most reliable benchmarks I have ever encountered for judging the character of someone has been: how they treat their furry frens. People can be consistent fuck ups in just about every other way, but if they take good care of their fren, I'm gonna think "jury still out on that person".
What’s the smartest move on Schumer’s leadership right now?
1️⃣ Start a transition to a new leader
2️⃣ Hold a leadership vote now
3️⃣ Keep him, but set clear benchmarks
Vote below! 📊
No register needed👇🏻
Vote below! Results in 24h 📊
1️⃣ Start a transition to a new leader
2️⃣ Hold a leadership vote now
3️⃣ Keep him, but set clear benchmarks
Vote below! 📊
No register needed👇🏻
Vote below! Results in 24h 📊
What’s the smartest move on Schumer’s leadership right now?
1️⃣ Start a transition to a new leader
2️⃣ Hold a leadership vote now
3️⃣ Keep him, but set clear benchmarks
Vote below! 📊
No register needed👇🏻
Vote below! Results in 24h 📊
www.getqibble.com
November 12, 2025 at 4:14 AM
What’s the smartest move on Schumer’s leadership right now?
1️⃣ Start a transition to a new leader
2️⃣ Hold a leadership vote now
3️⃣ Keep him, but set clear benchmarks
Vote below! 📊
No register needed👇🏻
Vote below! Results in 24h 📊
1️⃣ Start a transition to a new leader
2️⃣ Hold a leadership vote now
3️⃣ Keep him, but set clear benchmarks
Vote below! 📊
No register needed👇🏻
Vote below! Results in 24h 📊
What’s the smartest move on Schumer’s leadership right now?
1️⃣ Start a transition to a new leader
2️⃣ Hold a leadership vote now
3️⃣ Keep him, but set clear benchmarks
Vote below! 📊
No register needed👇🏻
Vote below! Results in 24h 📊
1️⃣ Start a transition to a new leader
2️⃣ Hold a leadership vote now
3️⃣ Keep him, but set clear benchmarks
Vote below! 📊
No register needed👇🏻
Vote below! Results in 24h 📊
What’s the smartest move on Schumer’s leadership right now?
1️⃣ Start a transition to a new leader
2️⃣ Hold a leadership vote now
3️⃣ Keep him, but set clear benchmarks
Vote below! 📊
No register needed👇🏻
Vote below! Results in 24h 📊
www.getqibble.com
November 12, 2025 at 4:14 AM
What’s the smartest move on Schumer’s leadership right now?
1️⃣ Start a transition to a new leader
2️⃣ Hold a leadership vote now
3️⃣ Keep him, but set clear benchmarks
Vote below! 📊
No register needed👇🏻
Vote below! Results in 24h 📊
1️⃣ Start a transition to a new leader
2️⃣ Hold a leadership vote now
3️⃣ Keep him, but set clear benchmarks
Vote below! 📊
No register needed👇🏻
Vote below! Results in 24h 📊
What’s the smartest move on Schumer’s leadership right now?
1️⃣ Start a transition to a new leader
2️⃣ Hold a leadership vote now
3️⃣ Keep him, but set clear benchmarks
Vote below! 📊
No register needed👇🏻
Vote below! Results in 24h 📊
1️⃣ Start a transition to a new leader
2️⃣ Hold a leadership vote now
3️⃣ Keep him, but set clear benchmarks
Vote below! 📊
No register needed👇🏻
Vote below! Results in 24h 📊
What’s the smartest move on Schumer’s leadership right now?
1️⃣ Start a transition to a new leader
2️⃣ Hold a leadership vote now
3️⃣ Keep him, but set clear benchmarks
Vote below! 📊
No register needed👇🏻
Vote below! Results in 24h 📊
www.getqibble.com
November 12, 2025 at 4:14 AM
What’s the smartest move on Schumer’s leadership right now?
1️⃣ Start a transition to a new leader
2️⃣ Hold a leadership vote now
3️⃣ Keep him, but set clear benchmarks
Vote below! 📊
No register needed👇🏻
Vote below! Results in 24h 📊
1️⃣ Start a transition to a new leader
2️⃣ Hold a leadership vote now
3️⃣ Keep him, but set clear benchmarks
Vote below! 📊
No register needed👇🏻
Vote below! Results in 24h 📊