talent ops + knowledge systems. librarian building the exec talent context layer @microsoft. prev @stripe @meta. hire me
Chicago, IL
Joined April 2008
- Tweets3.9K
- Following1.3K
- Followers476
- Likes5.5K
all the vets around me are now PE-backed. calls, texts, and emails are all answered by a shitty AI who can only help me book appts or ask me how likely i am to refer friends/family to them. get me and my dog out of this hellscape
This is so cool! I added my name (and my family members) to NASA's Dragonfly....Titan here we come!!!!
Your next trip: a flight to another world. 🪐
Send your name aboard NASA’s Dragonfly mission to Saturn’s largest moon Titan and get a personalized boarding pass for the journey.
Become a Dragonflyer today: go.nasa.gov/dragonfly-names
ALT Poster with a peach-and-gold background that reads “Journey to Titan” in bold orange text, with a banner beneath it that says “Fly your name to another world!” Below the headline, NASA’s Dragonfly rotorcraft flies through a hazy sky with its rotors spinning, with faint hexagon patterns in the background.
you'd think it natural that claude would have its own recurring human protagonist going toe-to-toe with @thsottiaux and @poteto
1. is it weird i find it cute that LLMs get smarter and faster when we give them time to sleep and dream?
2. love the aesthetic here. can't wait until vibecode design moves back to information dense visuals like these and fewer sub-sub headers/capsules.
Today on the blog, learn about our new instance of a “sleep” phase in continual learning, which enables LLMs to continuously integrate new facts and enhance reasoning while mitigating catastrophic forgetting or static knowledge cutoffs.
More →goo.gle/4iICkyD
ALT Overview of our sleep framework: Input streams from the active phase are stored in short-term memory before entering the sleep cycle, which bifurcates into Consolidation (transferring fragile short-term tokens into expanded long-term parametric memory using low-rank structures and MoE) and Dreaming (autonomous self-improvement via synthetic routing with random experts and RL refinement).
aaaand this led me to learning the history of the Google San Flex font from this gorgeous site... design.google/library/google…
i guess i need to get used to things called like "exponential time hypothesis" being... not made up shit
Elizabeth Below retweeted
If you've noticed a distinct dialect emerging in frontier AI models and struggled to understand what they're talking about, you're not crazy!
We turned to psycholinguistics to investigate what makes the writing style so weird, and why this might be happening
🚨New preprint! 🧵
while i have learned so much this year, my drive to read and ask questions and understand has mostly rewarded me with dread and anxiety.
not this.
interpretability might be the most important scientific problem in the world right now, as I don't think we can solve alignment if we don't know what we're building. I think I have good news about it. I co-authored early work in mechanistic interpretability at OpenAI and came away pessimistic, because neural networks appeared to fold into arbitrary tangled knots. years later, I started openai's first large-scale automated interpretability project, finding about 2% of neurons legible with the primitive tools we had
with modern agentic tools and a much closer look, the picture looks totally different. the vast majority in models up to trillions of parameters are legible, and their circuits beautiful. many famous examples of tangled neurons turned out to be misunderstood. The roadblocks I expected mostly aren't there. I now think ambitious interpretability will be solved by the time we need it most, and possibly quite soon, as automated scientists keep improving. it may be a bitter lesson for interpretability: the simple thing, at absurd scale, works
already-clean.goodfire.com/
Elizabeth Below retweeted
I'll say it again for the AI risk experts in the back:
We've already run a real life, large scale experiment on what happens when a superintelligence is unleashed on a narrow objective.
Look no further than U.S. capital markets.
Millions of investors, analysts, traders and algorithms, backed by trillions of dollars and wired together through prices, all chasing one goal: the highest risk-adjusted return on capital.
And it is spectacularly good at its job. It funds the best ideas, punishes the worst ones, and turns new information into prices faster than any of us can react. No individual can out-think it. 99+ percent of professionals can't beat it.
But look at what else it's done along the way:
- Optimized the measure instead of the intent. Quarterly earnings became the scoreboard, so companies learned to manage the scoreboard.
- Treated anything without a price as free. Pollution, addiction, obesity, depression, political fragmentation, fragile supply chains, the long-term health of the communities it operates in.
- Found every gap in the rules, then paid lobbyists to help write the next set and pull up the ladder behind incumbents.
- Remade the world around its objective. Companies, pension funds and our retirement accounts now answer to the S&P 500.
- Made itself impossible to switch off, because everything else now depends on it.
None of this required malice. Nobody sat in a room and chose these outcomes. It's simply what a very capable optimizer does when its objective is a narrow proxy for what we actually care about.
And note that we never solved this alignment problem. We've managed it, imperfectly, with regulation, taxes, disclosure rules and the occasional crisis to show us where the guardrails were missing. Over time, the market has captured the instutions designed to constrain it. We've been at it the better part of a century, and the misalignment is the worst it's ever been.
So the next time someone tells you AI is only dangerous if it "wants" something bad, or if its goals are mis-specified, point them at the S&P 500. It doesn't want anything. It just optimizes.
That's the whole problem.
I guess this rebrand was inevitable but corporate librarians and knowledge mgrs have been doing this work for 20 yrs
i’m finding my ai use has been making my own personal reliance on timestamps way harder… maybe just the influx and speed of information being created, so Creation Date, Last Modified/Opened, etc – these logs just don’t help as much as they used to.
Elizabeth Below retweeted
You can request Chicago here! graph.civlab.org/request?j=c…
Elizabeth Below retweeted
I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.
this is so wholesome (and PLUR) but also... something about how this guy tells the story and how the video is edited is so *SOLID*, like refreshing in some non-tiktok way?
Meet Josiah: he became a GRiZ fan a month ago, went to his first live show at Red Rocks — and somehow helped save the night for 10,000 fans by having his travel sax in the car because he just likes to “jam a little.”
He doesn’t have social media and likes it that way, but after some searching, I was able to track him down for an interview to hear exactly how it all went down. 👇🎷