jkobject Profile Banner
Jérémie Kalfon Profile
Jérémie Kalfon

@jkobject

Followers
503
Following
3K
Media
87
Statuses
2K

Doing a Ph.D. AI in Bio. | Ex @WhiteLabGx @BroadInstitute @MIT | Built @PiPleteam | ML, Cancer, Genomics, Data Sci, Entrepreneur, FullStack Dev | All Views Mine

Paris, France | Cambridge, MA
Joined May 2012
Don't wanna be here? Send us removal request.
@GabLattanzio
Gabriel Lattanzio
4 days
Il faut comprendre ce qu'a fait le PS dans la séquence, et ce que soutiendront le RN d'une part et le reste de la gauche d'autre part (même s'ils demandent davantage d'efforts aux enfants encore). Ils ont obtenu qu'une toute petite génération, née en 1964, fasse monter
@BFMTV
BFMTV
4 days
Retraites: Sébastien Chenu (RN), déclare que le parti "votera pour" la suspension de la réforme examinée à l'Assemblée nationale demain
35
242
924
@jkobject
Jérémie Kalfon
8 days
Without double-talk and with amazing panelists🧑‍🔬: - Yann Fleureau, CEO, Blossom Life Sci & Founder of Cardiologs - Steven Jerome, Director, Lead of Hit Discovery, Schrödinger - Jérémy Besnard, Advisor, InFocusTx & Co-founder of Exsciencia - Sofia Dahoune, Partner at Daphni 2/3
1
0
0
@jkobject
Jérémie Kalfon
8 days
🌐🧬I am excited to present you a round table I am doing together with Matteo Marengo Gabriel Michaux as part of our emerging Nucleate Parisian chapter led by Clara Brouaux 🔥. Title: **Inside AI: Choosing the Right Path to Value Creation** 1/3
1
0
0
@patrickc
Patrick Collison
12 days
An interesting trend we're noticing at Stripe: US startups are pulling ahead of their peers elsewhere. These charts show averaged revenue growth for software startups in each location. US startups typically grow somewhat faster than those elsewhere. However, since mid-2023, US
226
394
3K
@KaplanBen_Fr
Kâplan
13 days
Ce raisonnement de défendre l'indexation des retraites sur l'inflation est d'une malhonnêteté crasse. On n'a pas entendu Philippe Juvin défendre l’indexation du salaire des actifs, des fonctionnaires par exemple, sur l'inflation. C'est terrifiant de malhonnêteté. 😳 Résultat :
@LCI
LCI
13 days
🗣 Gel des retraites: "Nos marges de manoeuvre sont sur la baisse de la dépense. Il faut arrêter de taxer les Français." ➡ @philippejuvin invité de LCI
31
63
315
@vonderleyen
Ursula von der Leyen
19 days
Touchdown in Stockholm and working dinner with @SwedishPM and Swedish companies. Europe is working hard on making more private capital available, so entrepreneurs can scale up in Europe. We are building a Savings and Investment Union learning from the positive experience of
654
258
2K
@jkobject
Jérémie Kalfon
19 days
It was really tough to make Grok very lightly agree that sex isn't binary. Clearly got stuck on a logical loop, binary vs continuous, observation vs definition. It is terrifying to me that we have already created really powerful AIs and used them to spread disinformation..
@grok
Grok
19 days
@jkobject @oroyoroyoroy @elonmusk In ovotesticular DSD, biological sex isn't binary in every phenotypic sense, but individuals are classified male or female based on predominant gonadal function, external genitalia, or chromosomal makeup—e.g., 46,XX cases lean female if ovarian tissue dominates. Why? Sex
0
0
0
@sc_cath
Sylvain Catherine
19 days
Le patrimoine de tous les français a cru en pourcentage du PIB. C'est ce qu'il se passe lorsque le prix des actifs croit plus rapidement que la production. D'après la World Income Inequality database, la part du patrimoine net détenu par le top 1% est passé de 24,7% en 1996 à
@LCI
LCI
19 days
🗣 "On a des grandes fortunes qui sont passées en 1996 à 6% du PIB en patrimoine à 46% du PIB en presque 30 ans. Le poids des très grandes fortunes a explosé." ➡ @gabriel_zucman sur LCI
44
142
622
@DrSamuelBHume
Samuel Hume
20 days
This is genuinely paradigm-shifting 1-3% of patients diagnosed with (relatively) common diseases – atopic dermatitis, inflammatory bowel disease, and multiple sclerosis – actually have much rarer monogenic conditions, revealed by genome sequencing Given the obvious importance
@cremieuxrecueil
Crémieux
21 days
I just read an incredible paper. These authors went and used sequencing data from the U.K. Biobank and found that a sizable number of people were likely misdiagnosed with a common disease when they really have a rare one. The next part was amazing: They looked at sequencing
20
160
1K
@rohanpaul_ai
Rohan Paul
21 days
💊 Incredible. A new AI model, DrugReflector, that uses gene activity data from human cells to predict useful drug compounds much faster than brute-force screening. up to 17x better hit-finding than random library picks, and after feeding back its first test results the model’s
7
49
257
@demishassabis
Demis Hassabis
21 days
Very excited about our progress on materials! Super cool work, come join the AI for Science team.
@simonbatzner
Simon Batzner
23 days
Our team at DeepMind is growing (again). 🚀 We're tackling grand challenges in semiconductors, magnets, energy materials, superconductors, and beyond. Join us! Two positions below.
65
152
2K
@jkobject
Jérémie Kalfon
21 days
Lots of hype! As of now, in Arc's own Virtual Cell Challenge, over the 1000 teams and 10,000 submissions, only 1 managed to do better than nothing on all three metrics.
@VirtualCellNews
VirtualCell
25 days
And there have been significant recent advances in models that "unlock some key functionality of human cells’ workings that wasn’t available before." This includes State -- the first virtual cell model released by the Arc Institute – which  measures how sets of cells move in the
0
0
4
@sc_cath
Sylvain Catherine
23 days
C’est de la répartition : tout est transfert de pouvoir d’achat entre les uns et les autres, avec une perte sèche de richesse causée par la distorsion fiscale des cotisations. Faire ressentir immédiatement aux uns les cadeaux octroyés aux autres est peut-être la meilleure façon
@Clemence_Guette
Clémence Guetté
23 days
Baisser les pensions de retraites tous les ans. Une immense perte de pouvoir d'achat pour des millions de personnes, à commencer par les plus pauvres. Voilà ce qu'a prévu Lecornu pour faire payer aux Français son petit décalage de 3 mois de la réforme des retraites. Une
15
117
513
@malikules
Malika 🧬
25 days
private equity is really great and all but i will admit that i think the culture of biotechies and scientists is really interesting and i sometimes wish that i had the necessary physiology to work in them. at least for a bit. their entire life seems to be just one long sleepless
@owl_posting
owl
25 days
biology is really great and all but i will admit that i think the culture of ib/pe and big law is really interesting and i sometimes wish that i had the necessary physiology to work in them. at least for a bit. their entire life seems to be just one long sleepless night to finish
14
14
252
@jkobject
Jérémie Kalfon
25 days
It should fall a lot more
@zephyr_z9
Zephyr
25 days
BRUH Google lost over $100B in Market cap due to this tweet
0
0
0
@Valen10Francois
François Valentin
25 days
In a predictably 🇫🇷 twist, unions are blaming the Louvre fiasco on lack of staff They claim that 100s of jobs were cut in the last 10/15 years. The French press took these number for granted So I did what no one did and looked at the annual reports: The unions are wrong
27
277
1K
@sc_cath
Sylvain Catherine
27 days
Personnellement, je n’ai pas d’opposition de principe à la taxation de l’héritage ou du patrimoine immobilier — du moins pas plus qu’à celle du travail. Et si l’on pouvait, de manière crédible, augmenter la taxation des héritages afin de réduire celle du travail, j’y serais
54
67
415
@jkobject
Jérémie Kalfon
29 days
Because I saw it retweeted so many times, the truth seems to be that -Renewables didn’t cause Spain blackout- who said otherwise to misled? why? https://t.co/7CKmv3ZWCl
Tweet card summary image
npr.org
When millions lost power in Spain and Portugal this spring, some were quick to blame too much solar and wind power. That wasn't the cause, but the misinformation had an impact.
0
0
0
@jkobject
Jérémie Kalfon
29 days
"The experimental feedback loop kills you"
@tsuname
pablo
1 month
Just read the paper, and first of all, it's a good paper! They actually went and did pre-training all the way to downstream post-training. The models are small but compared to trillion token LLMs, single cell data is just not that big. Retaining model arch was smart. BUT...here's
0
0
1