Deadlifts and proof-of-work are essential to life | ETH Zurich
Zurich, Switzerland
Joined March 2009
- Tweets5.1K
- Following498
- Followers1.3K
- Likes7.9K
Pinned Tweet
Very excited to share our latest paper!
We show that the oxygen sensor PHD1 controls muscle mTORC1 in a hydroxylation independent manner by stabilizing the leucine sensor Leucyl tRNA Synthetase (LRS). Published today in @natcomm (a thread, 1/15).
nature.com/articles/s41467-0…
Every uni lab should have a dedicated person whose job is to increase efficiency using AI tools.
Data analysis, automation, writing…
Institutions should provide enterprise-level AI solutions for everyone.
What used to be the holy grail of knowledge is falling behind, and fast
85
Tip for ‘pre’ peer reviewing your own manuscripts.
Make an AI Agent generate a 30 page deepdive on your favorite three researchers. Let it skim podcast, papers, blogs… anything written.
1/3
1
153
Upload that doc into a custom GPT a call it your personal board of peer reviewers.
Now upload your manuscript into the board of reviewers and let it ask critical questions.
2/3
1
91
Gommaar D'Hulst retweeted
Alain-Laurent Verbeke, hoogleraar aan Harvard, vindt dat de slinger aan de universiteit te hard is doorgeslagen naar woke:
"Er is een beweging van zeer intolerant, radicaal, links gedachtegoed dat niet kan aanvaarden dat ook andere meningen bestaan"
#deafspraak
This week, I got my fifth tenure-track rejection in the last two years. All close calls.
Science is awesome, but academia sucks so damn hard.
4
474
Gommaar D'Hulst retweeted
Testing out DeepSeek-R1's ability to do "science" in one-shot.
I asked it to critique my 2017 R01 proposal which was scored Impact 26, 10% percentile, and funded by NINDS. (I wrote the proposal, I have the rights to share it with anyone or anything).
I uploaded a PDF of my Specific Aims and Research Strategy with the following, off-the-cuff prompt. It spit out a credible critique indistinguishable from a median quality reviewer's submission after 10 seconds of thought. Overall impact score 2. Thank you!
If this is what a free tool can do TODAY in 10 seconds of inference, on the first shot with zero iteration in prompting or followup, what actual role are humans going to play in the process of science going forward?
Would you rather have your multi-million dollar, career milestone proposal be critiqued by a super-intelligent AI agent, or by a tired, over-worked, potentially biased human?
It seems quite plausible that generating new, evidence-supported hypotheses and experimental plans could already be better accomplished by frontier models (with or perhaps without human prompting) than by humans alone.
What do we do now?
TLDR - mTOR is your friend. Not your enemy 😀
On September 28th, I decided to stop rapamycin, ending almost 5 years of experimentation with this molecule for its longevity potential.
I have tested various rapamycin protocols including weekly (5, 6, and 10 mg dose schedules), biweekly (13 mg) and alternating weekly (6/13mg) to optimize rejuvenation and limit side effects.
Despite the immense potential from pre-clinical trials, my team and I came to the conclusion that the benefits of lifelong dosing of Rapamycin do not justify the hefty side-effects (intermittent skin/soft tissue infections, lipid abnormalities, glucose elevations, and increased resting heart rate). With no other underlying causes identified, we suspected Rapamycin, and since dosage adjustments had no effect, we decided to discontinue it entirely.
Preclinical and clinical research has indicated that prolonged rapamycin use can disrupt lipid metabolism and profiles [1], as well as induce insulin and glucose intolerance [2] as well as pancreatic Beta-cells toxicity [3]. Despite anecdotal evidence of rapamycin slowing down tumor growth, its effect in inhibiting natural killer cells [4] do raise concern for anti-cancer immune surveillance and cancer risk in the longer run.
Additionally, on October 27th, a new pre-print [5] indicated that Rapamycin was one of a handful of supposed longevity interventions to cause an increase/acceleration of aging in humans across 16 epigenetic aging clocks. This type of evaluation is the first of its kind, as most longevity interventions up to date have been tested against one or two aging clocks, leading to invisible biases and potential intended “cherry picking” of favorable clocks for the tested interventions.
Longevity research around these experimental compounds is constantly evolving, necessitating ongoing, close observation of the research and my biomarkers which my team and I do constantly.
Sources:
[1] pubmed.ncbi.nlm.nih.gov/1217…
[2]pmc.ncbi.nlm.nih.gov/article….
[3]diabetesjournals.org/diabete…
[4]pmc.ncbi.nlm.nih.gov/article….
2
385
Gommaar D'Hulst retweeted
The hallmark of expertise is no longer how much you know. It's how well you synthesize.
Information scarcity rewarded knowledge acquisition. Information abundance requires pattern recognition.
It's not enough to collect facts. The future belongs to those who connect dots.
Studying something with the ‘latest technologies’ and drowning the reader with figures and tables has gotten priority over good research ideas.
1
121
