Evan Frick’s paper in this week’s science, here. What’s really interesting to me is not so much the result (which isn’t really surprising: that terrestrial mammal food webs have been collapsing) but the use of AI to address the complexity of how the ‘phenotype’ of the biosphere changes over time.

Our newly minted center at George Mason’s Schar School, here. With my co-director, Naoru Koizumi, we are building a new nexus between the clinic, the bench and policy makers.

Loyal readers know that I am fascinated with the red planet. Hat tip to Tyler on this blogpost from Casey Handmer. Casey worked at Hyperloop and NASA’s Jet Propulsion Laboratory before founding his own company. A Caltech PhD, he knows his stuff.

The story from SCIENCE, is here. The upshot is that journal articles that result from US-funded research have to be made available for free to the public as soon as published. No more year-long paywall. The idea, is that the taxpayer already paid for the work once and that having to pay again for access is double-billing. Apparently this also goes for the data. I’m for this, but how this will actually be implemented isn’t clear. Someone has to pay for the editorial labor. Will the publishing fees now be written into a grant proposal budget as allowable direct costs? A lot of business models are going to have to change I think.

One caveat: all of this is reversible by a successor Administration.

• What happens at the singularity in a Black Hole
• Why the parameterization of the universe (e.g. Planck’s constant) seem tuned for us to exist
• What in the brain explains the ‘hard problem’ of consciousness
• Why plant phospholipid metabolism is so different from animals or yeast
• The proto-function of the ribosome before proteins came on the scene
• Before the Big Bang
• Fermi’s paradox
• When will the next pandemic hit
• Will humans persist into the 22nd century

It is being deployed across the entire waterfront of science from particle physics to ecology. We use it to replace routine human labor, adapt models, direct autonomous vehicles and to reconfigure sensors in response to interesting events. There is still a human in the loop, but one can see the not so distant future where the artifact will be creating the hypothesis and testing it experimentally.

From the performance of the most recent Large Language Models, it’s not such a far jump to imagine accurate Wikipedia entries by AI authors–although, there is always the alignment problem: what is to stop the AI from producing realistic-looking drivel, either accidentally or to fool us.

Sad to report the passing of my former deputy at NSF, Dr. Jane Silverthorne, a superb plant biologist.

As part of your short-list visit to campus, you can expect to be hosted for dinner usually at a nice restaurant (although it could be at a faculty member’s house). This will come at the end of what has already been a long day. Don’t relax. This dinner will be all work and no play. From a practical standpoint, it would be good if you have time to clean up a bit after your long day. If you brought something else to wear that is appropriate to the occasion, then changing into those clothes would be helpful. If you order, try to order food for which there is no chance of it spilling on to you or others (red Italian sauces should be off the list). Aim for a price point that is in the middle of the menu and that doesn’t stand out. If you have special dietary needs, those should he communicated to the host before you travel. You may be offered an adult beverage (i.e. alcohol) either before or during dinner. Politely decline it. The offer is, among other things, a test of your judgement for which there is only one correct answer. You will need to be completely sober during the dinner. Note that the other attendees may well imbibe a bit.

The dinner will be a continuation of the interview process. It may masquerade as convivial conversation, but it is not. Each opportunity for you to converse will be analyzed by the attendees for the key factors of: intelligence, fitting in and good judgement. All of these are considered to be necessary for the success in the search. Do not discuss politics or volunteer your political opinions. Do not monopolize the conversation. Try to actively listen to each person as they speak. Make eye contact with the other attendees as they are speaking. Often, it’s a good idea to feed their question back to them as a question in order to buy time to think and to make certain that you both are on the same page.

At the end of the evening, make certain to explicitly thank the host and the other attendees for their time and hospitality. And you aren’t off camera until you are back in your hotel room with the door closed. Don’t even think about going out afterwards.