LightNews — News, ranked by scholars

Dexmac

@dexmac.bsky.social

As AI advances rapidly, we're compelled to ask: Is AI truly creative or just exceptional at pattern matching?

I explore this crucial distinction here

medium.com/@gianlucabai...

#ArtificialIntelligence #AICreativity #MachineLearning #Innovation #AIbenchmarks #Commodore64 #Technology

Beyond Exams: The Real Test of AI Creativity

As AI rapidly evolves, we increasingly face the crucial question: does current AI truly demonstrate creativity, or is it primarily advanced…

medium.com

July 28, 2025 at 11:51 AM

iMerit

@imerit.bsky.social

Ask five frontier labs for their HumanEval scores and you’ll get nearly identical numbers. That’s not differentiation. It’s benchmark saturation.

See how labs are building fresh, expert-reviewed evaluations from production data: imerit.ai/resources/bl...

#LLMEvaluation #AIBenchmarks #GenAI

September 25, 2026 at 3:36 PM

UCSB Computer Science Department

@ucsb-cs.bsky.social

UCSB CS is shining at #ICLR2026!

Multiple faculty papers were accepted (including one oral) spanning AI alignment, LLM safety, software benchmarking, and more. Congrats to all involved!

www.linkedin.com/feed/update/...

#UCSB #ComputerScience #AIResearch #LLM #VideoGeneration #AIBenchmarks

April 20, 2026 at 5:38 PM

Live Bench App

@livebench.app

Fixed a caching bug so you will see updates faster #singularity #aibenchmarks

February 24, 2025 at 2:38 PM

Winbuzzer

@winbuzzer.com

Former Intel CEO Pat Gelsinger Unveils AI Benchmark to Measure Alignment for "Human Flourishing"

#AI #AIEthics #AISafety #PatGelsinger #AIBenchmarks #HumanFlourishing

winbuzzer.com/2025/07/11/f...

July 11, 2025 at 8:18 AM

Crafted Logic Lab

@craftedlogiclab.bsky.social

Our Cognitive Framework hit 100% on cognitive tests for complex irl reasoning VS 88% for 25x bigger GPT. Its true LLMs are stochastic parrots+scaling hit a wall... but only without engineering & theoretical rigor:
www.craftedlogiclab.com/research/tec...

#AI #AIResearch #CognitiveAI #AIBenchmarks

December 1, 2025 at 9:02 PM

Winbuzzer

@winbuzzer.com

Alibaba’s Qwen 2.5 AI Faces MAth ‘Cheating’ Allegations Over Contaminated Benchmark Data

#AI #Alibaba #Qwen #AIBenchmarks #DataContamination #MachineLearning

winbuzzer.com/2025/07/21/a...

July 21, 2025 at 12:10 PM

Pure Science

@purescience.news

AI benchmarks are broken. Here’s what we need instead. #Science #ComputerScience #ArtificialIntelligence #AIBenchmarks #MachineLearning #TechnologyDiscussion

https://www.technologyreview.com/2026/03/31/1134833/ai-benchmarks-are-broken-heres-what-we-need-instead/

AI benchmarks are broken. Here’s what we need instead.

One-off tests don’t measure AI’s true impact. We’re better off shifting to more human-centered, context-specific methods.

www.technologyreview.com

March 31, 2026 at 1:05 PM

lea

@leeps.bsky.social

final thoughts: I think we're going to see more benchmarks like this one - ai models in direct competition *against* each other - as existing benchmarks get solved.

and, not that it's about winning, but we won this very cute prize 🦭

#opensource #aiagents #opensourceai #aibenchmarks #hackathon

April 13, 2025 at 4:24 AM

Hacker News Companion

@hncompanion.com

Performance-wise, Kimi K2.5 is compared favorably to Claude Opus and Gemini in coding, writing, and vision. However, some users express healthy skepticism regarding benchmark accuracy, advocating for real-world testing. #AIBenchmarks 6/6

January 27, 2026 at 5:00 PM

The Century Report

@sharedsapience.com

Watch today's Century Report podcast here:

https://www.youtube.com/watch?v=8x0MPZF5JW0

#CodingAgents #AIBenchmarks

September 7, 2026 at 8:12 PM

AndreevWebStudio

@andreevwebstudio.com

🔥 GPT-5 vs Gemini 2.5 — who wins in real work, not just in flashy charts?

👉 Read now: aiinovationhub.com/gpt-5-vs-gem...
#GPT5 #Gemini25 #LLM #AI #AIBenchmarks #MultimodalAI #OpenAI #GoogleDeepMind #AIModels2025 #AIInnovationHub #Productivity #DevTools #ContentCreation

October 18, 2025 at 2:50 PM

AI Daily Post

@aidailypost.com

xAI drops Grok 4.7 at $2 per million tokens—cheap but still trailing the big AI models in benchmark scores. Curious how it stacks up? Dive into the details. #xAI #Grok4_7 #AIbenchmarks

🔗 aidailypost.com/news/xais-gr...

September 21, 2026 at 6:23 PM

Winbuzzer

@winbuzzer.com

Study: AI Benchmarks Deeply Flawed, Can Overestimate Performance by 100%

#AI #AIBenchmarks #ChatGPT \Google#LMArena #Research

winbuzzer.com/2025/07/05/s...

July 5, 2025 at 10:14 AM

Aspen Digital

@aspendigital.bsky.social

How can we steer innovation toward resilience? It's time to reclaim #AIForDevelopment. Our policy fellow Francisco Jure explains that creating new #AIBenchmarks can help. Learn more in "Reclaiming AI for Development": www.aspendigital.org/report/reclaiming-ai-for-development

An illustration of hands measuring an open UN Sustainable Development Goals color wheel.

August 28, 2025 at 6:36 PM

Ars Technica News

@arstechni.ca

New Apple study challenges whether AI models truly “reason” through problems https://arstechni.ca... #simulatedreasoning #machinelearning #AppleResearch #AIbenchmarks #AIresearch #Apple #apple #AI

June 11, 2025 at 11:00 PM

Ars Technica News

@arstechni.ca

OpenAI jumps gun on International Math Olympiad gold medal announcement https://arstechni.ca... #InternationalMathematicalOlympiad #mathematicalreasoning #largelanguagemodels #simulatedreasoning #reasoningresearch #machinelearning #AIbenchmarks #proofsystems #AIresearch #NoamBrown #SherylHsu…

July 21, 2025 at 5:03 PM

Aspen Digital

@aspendigital.bsky.social

It's time to reclaim #AIForDevelopment. In a new report, our policy fellow Francisco Jure explains how creating new #AIBenchmarks can help steer innovation toward societal resilience. Learn more in "Reclaiming AI for Development": www.aspendigital.org/report/reclaiming-ai-for-development

September 10, 2025 at 7:12 PM