@caliber_meter

Enhancing human potential in the AI era. Verify real capability in an AI-assisted world. https://xitter.cf/t.co/IKQbqUHpnX

 California, USA
caliber.dimaggi.ai
Joined December 2024
  • Tweets91
  • Following47
  • Followers14
11Media
Caliber retweeted
Yann LeCun
@ylecunSep 21
Replying to @Christiano92 @musedivision @geoffreyhinton
I have absolutely never ever "asked" to stop R&D on LLMs. In fact, I was very supportive of LLM efforts at Meta. Just look up my public advocacy for Galactica (November 2022), and the open sourcing of OPT-175B (May 2022), and of the Llama herd (mid 2023). I was actually influential in convincing the Meta leadership to start an AI-first product division built around Llama in January 2023. I've always said that LLMs were impressive, useful, and had amazing potential. I've also always said they were an off-ramp on the path towards human-level AI. Stop making me say things I didn't say or claim that I somehow hindered LLM R&D at Meta. That's BS.
52
43
3
792
79,661
This is the biggest challenge vis a vis the labs but recall that large companies have always had a huge resource advantage and yet sometimes small, highly motivated teams with incredible vision, focus, cohesion and insight blow away a big team. Welcome to capitalism 101.
Andrew Curran
@AndrewCurran_Sep 6
If you are competing against OpenAI or Anthropic, you are going up against people using models two generations ahead of the best public models in the world, burning through $7k–$10k worth of tokens per day on an infinite token budget, running at 8x normal speed.
6
3
1
28
4,062
Insane!
Alex
@The_AlexSep 5
Tell Astra to open Paint in the browser and try to draw you
42
Caliber retweeted
Peter Dedene
@dedeneSep 5
Me running my orchestra of agents from the terminal:
57
176
24
1,493
123,447
Turning in AI-assisted work is fine until someone asks you to defend the reasoning behind it. CALIBER - caliber.dimaggi.ai/plans evaluates evidence grounding and logic so students can back up their work, and educators can see who actually understands the material. Test your score with the Caliber - caliber.dimaggi.ai/plans #HigherEd #AIinEducation #Cognition #ArtificialIntelligence #AI #college #students #professor #highereducation #Education #humanpotential
🤖 Made with AI
2
4
63
Caliber retweeted
DIMAGGI AI
@dimaggi_aiJul 28
The Maintainer Loop: Keeping Books on the Bookkeeper dimaggi.ai/insights/ontology…

The Maintainer Loop: Keeping Books on the Bookkeeper

An agent maintains my LLM benchmark under one strict rule: it proposes, deterministic code decides, a human merges. Even then, I keep the schedules off.

dimaggi.ai
1
1
14

The Host-Side Ceiling on Agentic Inference Capacity

Article · @maggienanyonga
4
6
293
Verify real capability in an AI-assisted world. caliber.dimaggi.ai/meter
Brendan Nyhan (@BrendanNyhan on 🟦☁️)@BrendanNyhanJul 8
Friends don't let friends continue to give exams and papers that are not adapted to the current AI landscape (this is AFTER 27 bailed so the reality is even worse) PS Props to students 1, 22, and 31
1
203
Verify real capability in an AI-assisted world. caliber.dimaggi.ai/meter
🤖 Made with AI
Polymarket
@PolymarketJul 9
NEW: Ivy League professor suspected AI cheating, ordered an in-person final, & saw average scores plunge from 96 to 48.
6
1,836
Verify real capability in an AI-assisted world. caliber.dimaggi.ai/meter
Paul Graham
@paulgJul 9
A Brown professor gave his students a take-home midterm exam. After suspecting many cheated using AI, he made the final in-person. The orange dots are the midterm scores and the gray dots are the final scores. Looks like all but 3 cheated on the midterm.
1
125
Caliber retweeted
Caliber@caliber_meterJul 1
AI can polish a product memo, technical proposal, strategy brief, or essay. Can you defend the reasoning behind it? Caliber evaluates the evidence, trade-offs, and judgment behind AI-assisted work across professional and academic settings.

3
17
13,422
Widespread AI misuse means higher ed must rethink assessment news.cornell.edu/stories/202…

Widespread AI misuse means higher ed must rethink assessment | Cornell Chronicle

Large numbers of college students are now using artificial intelligence to complete – and cheat on – their assignments, suggesting that colleges and universities need to change how they are evaluat...

news.cornell.edu
31
Today, we’re launching our home robot Isaac 1. Isaac 1 deliveries will begin this fall. Order yours below.
1,607
1,129
1,995
16,685
18,795,010
Enhancing Human Potential... For most of working life, polish was a decent stand-in for thought. If the analysis read well, someone had probably done the reasoning, and that rough correlation is what let managers and teachers and clients judge quality at a glance. AI broke the correlation. The analysis reads well now whether or not any reasoning happened behind it. The work is as good, or as thin, as it ever was. What's gone is your ability to tell which, from the page in front of you. And you're still making the call on a cue that has gone silent. caliber.dimaggi.ai/meter
1
1
23
Measuring Judgment in an AI-Assisted World — Try Caliber caliber.dimaggi.ai/meter and share your summary reports as well as sharpen the defensibility score. Here's mine for a pet project that I'm working on.
Caliber@caliber_meterJun 23
AI can polish the work. Can you defend the reasoning behind it? Caliber measures judgment, evidence, trade-offs and uncertainty in AI-assisted work.

11
AI can polish the work. Can you defend the reasoning behind it? Caliber measures judgment, evidence, trade-offs and uncertainty in AI-assisted work.

2
3
2
43
180,916
Try caliber.dimaggi.ai and share your card reports.
2
222
There is a narrow window to embed structural security protocols before multi-agent systems scale globally. We believe this multilayered approach to agent security should be a collaborative priority for AI labs, government, and academia. See the framework → goo.gle/4vis97Q

Securing internal systems against increasingly capable and imperfectly aligned AI

Discover our AI Control Roadmap: a defense-in-depth system to securely manage advanced, potentially misaligned AI agents.

deepmind.google
9
21
4
109
33,518
Our data shows that the vast majority of issues don't stem from bad intent. They usually happen because an agent misinterprets a command or gets overly enthusiastic to achieve a goal. Understanding these nuances is critical for refining safety and security protocols. ⬇️
10
17
3
109
27,741
Instead of assuming AI will always do what we intend, we ask: what if it doesn't? That’s why we’ve developed our AI Control Roadmap: a framework for building and managing the advanced AI we deploy within Google. 🧵
89
93
33
593
93,388
Load more