reefecologist.bsky.social
postdoc on an island 🏝️
caribbean reefs 🐠 ecology 🌎 code 💻 stats 🧮 3D 📷 robots 🤖 AI 🧠 inverts 🪸
📍US Virgin Islands 🇻🇮🇺🇸
214 posts
1,163 followers
1,247 following
Regular Contributor
Active Commenter
comment in response to
post
Welcome to the 🦋!
comment in response to
post
Michael is jealous of this strategy, but also too shy to attempt it.
comment in response to
post
IMO these opportunities seem useful for democratizing these powerful resources. impactchallenge.withgoogle.com/genaiacceler...
comment in response to
post
Seems pretty cool. It’s the nonprofit arm of google. Good access to resources and funds for the cause. “Instead of just offering AI species suggestions of what you saw, we want to offer a why as well.”
comment in response to
post
Maybe a naive question : was this co-written by NSF and WH/OMB/DOGE?
comment in response to
post
Did you get the educational discount? slack.com/intl/en-gb/h...
comment in response to
post
Thanks for the article! The pi model and performance increase reminds me a lot of how toddlers and children learn in new environments (not something I know much about either, just an amateur observation). Super cool.
comment in response to
post
I see that. was more asking for future non- teleop vs.
comment in response to
post
Hello! Could you add me? Thanks!
comment in response to
post
Big thoughts and prayers energy here…
comment in response to
post
Fascinating! How does it know where things go?
Also my cat would lose it 😅
comment in response to
post
The IOB part was chilling. Human brains do some wild things on the brink of death.
comment in response to
post
omg doing that now. No hate to Eurovision. My American mind just cannot deal lol.
comment in response to
post
Omg Monroe looks a lot like our late Julep. She was also several updates behind 🥲
comment in response to
post
That’s good that you had sex balanced studies w/mice.
The paper I linked above did not, according to abstract: l
“Here we show that in male mice…”
I would hope human studies of any repute also have m/f balance.
But why only male mice for this fundamental stuff about sleep? And in Nature Comm!
comment in response to
post
I assumed things had changed in the year of our lord 2025.
Evidently NIH now enforces sex balanced designs (though who knows anymore)
All because female estrous cycles are too difficult to control for… The jokes write themselves.
comment in response to
post
Well I guess I’m oppressed 🥲
comment in response to
post
One day I was so bored during a flight delay that I looked it up. Not much meaning behind it. Just circles of fish.
I did chuckle at the title: “got any jacks”
www.donaldlipski.net/got-any-jacks
comment in response to
post
Wow! On the camera! It’s a scourge here in the Caribbean and has displaced most native seagrass. Do you guys know if FWC is doing anything to monitor its spread?
comment in response to
post
I’m just a lowly climate scientist trying to vouch for actual things that are harming the environment. here’s an NYU professor and RL expert making same appeal that LLMs are not destroying the planet, and citing the same source.
All the best.
bsky.app/profile/euge...
comment in response to
post
Will once again refer to the (cheat sheet version) of the substack article. It’s very helpful!
andymasley.substack.com/i/162196004/...
comment in response to
post
The substack article I posted included estimates for training into the per query estimate. It isn’t much.
I agree about other critiques of LLMs, and think it’s worth starting from that premise instead.
The climate impacts premise is a misinformed and cheap move that detracts from real issues.
comment in response to
post
Im in USVI.
Does registry = shipyard?
comment in response to
post
Evaporative cooling. Also most electricity is based on heating water to spin magnets around wires.
Even then, LLMs aren’t bad for the environment in terms of water use. And it dangerous red herring for other issues - sincerely an environmental scientist.
andymasley.substack.com/p/individual...
comment in response to
post
unfortunately those are usually the exclusives 😕
it's despicable to paywall this information in particular.
bsky.app/profile/reef...
comment in response to
post
Cool! How could they be using the DOM?
comment in response to
post
I don’t teach async but I can heavily relate. Love asking students to explain what this bit of their code does
comment in response to
post
I study sponges, but will pass on the message 🫡 🥲