Twitxy Saved

Taal Levi

Verified account This account is verified. Learn more

@taaltree

Wildlife ecology and conservation, metabarcoding, eDNA, disease ecology, mathematical modeling. Prof of Fisheries, Wildlife, and Conservation at Oregon State U
Corvallis, OR agsci-labs.oregonstate.edu/levit/2021/03/… Joined June 2009
1.6K Following 6.1K Followers 7.6K Posts

Latest posts 5

Taal Levi Verified account @taaltree
A few years ago I was given a large sum of money and asked to donate it to maximize impact. Because I am a wildlife scientist, I focused most giving in that space, but how to decide what maximizes impact? Here are some places I've donated to (~$650k now) and some thoughts for philanthropists / 1
Vince Buffalo @vsbuffalo Nov 7
I don’t know what I find more ridiculous: someone attacking donating money to *checks notes* mosquito nets, or the fact that they’re an academic that’s seemingly unconcerned with how terrible their argument is.
View quoted post on X
4 replies 3 reposts 33 likes 3.8K views View on X
Taal Levi Verified account @taaltree
If the blue whale population growth trajectory continues, we'll be back to the pre-whaling baseline when my daughter is my age. I'd love to see that!
New assessment of the greatest of all time: Antarctic blue whales, showing how whaling reduced them from >200,000 to a low of just 418 before slow recovery at 4.5% per year to more than 4000
Rand et al. [incl. me] doi.org/10.3354/esr015…
View quoted post on X
0 replies 3 reposts 17 likes 627 views View on X
Taal Levi Verified account @taaltree
This was almost definitely a bias report made by a graduate student teaching an online course, which is why they refer to both their supervisor and the department chair. Anyone can report anything they want, and it should surprise nobody that a graduate student instructor might be overzealous. It would reflect poorly on the university if the administration responded inappropriately, but we have no such information.
FIRE Verified account @TheFIREorg Nov 2
An Oregon State University instructor reportedly filed a bias report against a student for arguing that there are two biological sexes in a class assignment.

The university says its Bias Response Team doesn’t punish protected speech, but you don’t have to punish students to silence them. The prospect of being subjected to vague “restorative processes,” or having your views documented in a university file, is chilling enough.

And students are already reluctant to speak up. In FIRE’s College Free Speech Rankings survey, 52% of Oregon State students reported at least occasionally self-censoring with professors, while 41% expressed discomfort disagreeing with professors on controversial issues in written assignments.

Professors can grade students on their arguments, research, and understanding of course material. But reporting them for their personal views is another matter entirely, and teaches students a dangerous lesson: Keep your opinions to yourself.

@OregonState must scrap its Bias Response Team.
View quoted post on X
0 replies 0 reposts 7 likes 1.2K views View on X
Taal Levi Verified account @taaltree
Feel free to DM me issues that come up. I want this to be useful.
Taal Levi Verified account @taaltree Oct 22
Please share with the community. I'm looking for beta testers for CascadiaTools, an AI coding agent harness for ecological, wildlife, fisheries and conservation-genetic analysis. It is available as a @claudeai plugin. Our aims are to: (i) argue that the trust ecologists place in vetted R packages can be extended to analyses run by coding agents, if the agents call vetted code instead of writing it; (ii) state the design principles that make agent-produced analyses checkable; (iii) describe the architecture shared by all modules; (iv) report how each of 29 modules was built and independently verified, including the limits each verification exposed; (v) show how a researcher with no coding experience uses it, from question to publication figures; and (vi) outline how vetted libraries for coding agents could become a standard part of ecological practice. /1 cascadiawildlifelab.org/tools/index.ht…
View quoted post on X
Taal Levi Verified account @taaltree
Please share with the community. I'm looking for beta testers for CascadiaTools, an AI coding agent harness for ecological, wildlife, fisheries and conservation-genetic analysis. It is available as a @claudeai plugin. Our aims are to: (i) argue that the trust ecologists place in vetted R packages can be extended to analyses run by coding agents, if the agents call vetted code instead of writing it; (ii) state the design principles that make agent-produced analyses checkable; (iii) describe the architecture shared by all modules; (iv) report how each of 29 modules was built and independently verified, including the limits each verification exposed; (v) show how a researcher with no coding experience uses it, from question to publication figures; and (vi) outline how vetted libraries for coding agents could become a standard part of ecological practice. /1 cascadiawildlifelab.org/tools/index.ht…
cascadiawildlifelab.org CascadiaTools: validated ecological analysis for AI coding agents CascadiaTools lets an AI agent run wildlife, fisheries, occupancy, abundance, movement and conservation-genetic analyses with statistical code that has been checked by simulation and independently...