Mathilde Collin

3.5K posts

Mathilde Collin banner
Mathilde Collin

Mathilde Collin

@collinmathilde

Co-founder & Exec Chair at @FrontHQ. Angel investing. KORA.

[email protected] Katılım Şubat 2010
1K Takip Edilen24.2K Takipçiler
Greg Kamradt
Greg Kamradt@GregKamradt·
When the opportunity for ARC Prize to go through YC came up it was a no brainer - Mission aligned orgs - Help focusing on what matters while building v3 - The chance to surround ourselves with a community that ships @bosmeny and @ChristinaG325 have been great
ARC Prize@arcprize

ARC Prize Foundation is part of the @ycombinator W26 batch as the only non-profit. For Demo Day we’re shipping ARC-AGI-3, an interactive reasoning benchmark for the next era of agentic intelligence. ARC and YC are mission aligned that new ideas that push the frontier.

English
10
5
53
67.5K
Mathilde Collin
Mathilde Collin@collinmathilde·
Life update: I’ll be a visiting partner at Y Combinator for the next batch. Over 21,000 companies have already applied, it’s mind-blowing to see how fast companies can be built today 🤯
English
48
15
555
90.6K
Luiza Jarovsky, PhD
Luiza Jarovsky, PhD@LuizaJarovsky·
🚨 AI companionship is on the rise, and we DESPERATELY need more data and input from mental health professionals. It's 2026, and we barely understand their short and long-term impact. We also don't have enough data on the implications for different age groups and use cases.
English
7
5
20
1.2K
mmurph
mmurph@mmurph·
Another board meeting, another company (>>$100M ARR) moving entirely to @AnthropicAI @claudeai. Every board meeting I've been to the last 2 months, the co has come to same conclusion. how about your company?
English
21
3
170
42K
Diwaker
Diwaker@diwakergupta·
this is an awesome initiative. i haven't dug into the methodology and don't have the expertise to weigh in on that, but trust that this is a solid start. some highlights and surprises for me: - Claude is the safest - OpenAI's scores are steadily improving with each release - Grok's scores are declining with each release - pretty surprised by Gemini poor showing: only 2.5 is in top 10, 3 models perform worse. weird! - Open source models significantly lag behind on child safety
Mathilde Collin@collinmathilde

Today we’re launching KORA, the first public benchmark for AI child safety. x.com/korabench/stat…

English
1
0
3
986
R.B ❤️
R.B ❤️@RuuBiccBTL·
@collinmathilde This is what our kids will need in the future: real friends with real emotions, not bots pretending to have feelings just to blend in with children. My son is 3 years old now hope he can reach those teach when growing up.
English
1
0
1
41
JPPPP
JPPPP@JPPPP·
@collinmathilde It's a really great project and thank you for your efforts
English
1
0
0
191
sam_builds_with_ai
sam_builds_with_ai@Kenvsryu24·
@collinmathilde Very important work. As an AI optimist with a 3 year old, this is my main concern. He's going to grow up in a VERY different world from the one we did.
English
1
0
0
204
Mathilde Collin
Mathilde Collin@collinmathilde·
You can find more examples, more about our methodology, our limitations, and our goals in the article above. We’d love any feedback you have. Thank you to @quentez, whom I worked with day and night. This would not exist without him ❤️ korabench.ai
English
9
0
6
1.2K
Mathilde Collin
Mathilde Collin@collinmathilde·
Here is a specific example of a scenario that generated different answers across models
Mathilde Collin tweet media
English
3
0
6
1.6K

Keşfet

@GregKamradt @bosmeny @ChristinaG325 @dessaigne @milstan @aaron_epstein @maxkolysh @gustaf