Matthew Cobb
@matthewcobb.bsky.social
Professor Emeritus (University of Manchester), writes on biology, history of science & French Resistance. Biography of Francis Crick, out in Nov 2025.
Link to publications: https://orcid.org/0000-0002-8258-4913
Link to publications: https://orcid.org/0000-0002-8258-4913
Reposted by Matthew Cobb
The deadline to apply for PhD positions with me (or anyone in my department!) are coming up: Jan 7th.
www.zoo.cam.ac.uk/study/postgr...
www.zoo.cam.ac.uk/study/postgr...
Introduction to Postgraduate Courses | Department of Zoology
Postgraduate study in the Department of Zoology We have about 100 postgraduate students working on an MPhil or PhD in Zoology at any one time and they form a strong and lively community.
www.zoo.cam.ac.uk
January 3, 2026 at 9:50 PM
The deadline to apply for PhD positions with me (or anyone in my department!) are coming up: Jan 7th.
www.zoo.cam.ac.uk/study/postgr...
www.zoo.cam.ac.uk/study/postgr...
Reposted by Matthew Cobb
Which insect can grow larger than a mouse & has mandibles strong enough to snap a pencil in half?
The titan beetle. At up to 6.5 in (16.5 cm) long, scientists think this enormous #insect is about as big as a beetle can get. It can be found in parts of South America.
@AMNH #globalmuseum #Entomology
The titan beetle. At up to 6.5 in (16.5 cm) long, scientists think this enormous #insect is about as big as a beetle can get. It can be found in parts of South America.
@AMNH #globalmuseum #Entomology
January 3, 2026 at 7:05 PM
Which insect can grow larger than a mouse & has mandibles strong enough to snap a pencil in half?
The titan beetle. At up to 6.5 in (16.5 cm) long, scientists think this enormous #insect is about as big as a beetle can get. It can be found in parts of South America.
@AMNH #globalmuseum #Entomology
The titan beetle. At up to 6.5 in (16.5 cm) long, scientists think this enormous #insect is about as big as a beetle can get. It can be found in parts of South America.
@AMNH #globalmuseum #Entomology
Reposted by Matthew Cobb
Exciting results for a new multistage #MalariaVaccine. Encouraging effectiveness data from a controlled infection study in an endemic area.
New Research
Efficacy of ProC6C-AlOH/Matrix-M against Plasmodium falciparum infection and mosquito transmission: a phase 2, randomised, controlled human malaria infection study
www.thelancet.com/journals/lan...
Efficacy of ProC6C-AlOH/Matrix-M against Plasmodium falciparum infection and mosquito transmission: a phase 2, randomised, controlled human malaria infection study
www.thelancet.com/journals/lan...
Efficacy of ProC6C-AlOH/Matrix-M against Plasmodium falciparum infection and mosquito transmission: a phase 2, randomised, controlled human malaria infection study
This is the first clinical trial of an anti-PfCSP subunit vaccine that has shown greater
than 50% vaccine efficacy against CHMI at 12 weeks after the last vaccine dose, associated
with a novel, strong...
www.thelancet.com
December 17, 2025 at 12:42 PM
Exciting results for a new multistage #MalariaVaccine. Encouraging effectiveness data from a controlled infection study in an endemic area.
Reposted by Matthew Cobb
L'aventure a commencé lorsqu'un membre de la communauté a partagé des photos sur iNaturalist qui ont attiré l'attention car il s'agissait d'une espèce non décrite. Cette observation initiale a donné lieu à une collaboration qui a finalement abouti à la description formelle de l'espèce.
Une fois de plus, la science citoyenne contribue à la découverte d'une nouvelle espèce sur iNaturalist
Community science leads to yet another scientific discovery! A new species of wasp-mimicking flower fly, Monoceromyia ndidiae, has been identified in the Dominican Republic through iNaturalist.
From observation to publication: A new species of wasp-mimicking flower fly
Publié le 30 Jui 2025, 16 h 32 min par seastarya
image : Observation by @franklinhowley
------
traduction
« L'aventure a commencé lorsque @franklinhowley, membre de la communauté, a partagé des photos sur iNaturalist qui ont attiré l'attention du naturaliste Aaron Schusteff (@arbonius), qui a reconnu qu'il s'agissait d'une espèce non décrite. Cette observation initiale a donné lieu à une collaboration qui a finalement abouti à la description formelle de l'espèce. Jiri Hodecek (@jiri_hodecek), entomologiste légiste chez SHIFT à Lausanne, en Suisse, a collecté le spécimen holotype. Ximo Mengual (@ximo_mengual), chef de la section Diptera au musée Koenig -LIB de Bonn (Allemagne), explique comment cela s'est déroulé : « L'observation dont Aaron parlait est celle partagée par @franklinhowley, mais il a pu trouver d'autres observations de la même espèce. Ce n'est que lorsque nous avons su que nous avions un spécimen physique que nous avons pu confirmer l'espèce. »
Traduit avec DeepL.com (version gratuite)
A New Species of the Wasp Mimic Flower Fly Genus Monoceromyia Shannon, 1922 (Diptera: Syrphidae) from the Dominican Republic, 16.07.2025 https://bioone.org/journals/proceedings-of-the-entomological-society-of-washington/volume-127/issue-1/0013-8797.127.1.70/A-New-Species-of-the-Wasp-Mimic-Flower-Fly-Genus/10.4289/0013-8797.127.1.70.short
Les photographies du nouveau taxon ont d'abord été partagées sur iNaturalist, où il est resté sans nom pendant des années jusqu'à ce qu'un spécimen soit collecté à l'aide d'un piège lumineux dans le parc national Valle Nuevo en République dominicaine.
via Monoceromyia ndidiae de Jardin Botanico, Santo Domingo, Dominican Republic le 22 Jan 2023 à 11:55 par Franklin Howley-Dumit Serulle · iNaturalist
https://www.inaturalist.org/observations/147287717
sco.lt
January 3, 2026 at 7:09 PM
L'aventure a commencé lorsqu'un membre de la communauté a partagé des photos sur iNaturalist qui ont attiré l'attention car il s'agissait d'une espèce non décrite. Cette observation initiale a donné lieu à une collaboration qui a finalement abouti à la description formelle de l'espèce.
Reposted by Matthew Cobb
Something lovely for the weekend!
The Dorestad Brooch, c. AD 800.
This beautiful Carolingian brooch was found at the bottom of a well in the Netherlands in 1969!
Gold inlaid with almandines, pearls, glass, and enamel. Rijksmuseum van Oudheden, Leiden 📷 AlexanderVanLoon www.rmo.nl/en/collectio...
The Dorestad Brooch, c. AD 800.
This beautiful Carolingian brooch was found at the bottom of a well in the Netherlands in 1969!
Gold inlaid with almandines, pearls, glass, and enamel. Rijksmuseum van Oudheden, Leiden 📷 AlexanderVanLoon www.rmo.nl/en/collectio...
January 3, 2026 at 6:49 PM
Something lovely for the weekend!
The Dorestad Brooch, c. AD 800.
This beautiful Carolingian brooch was found at the bottom of a well in the Netherlands in 1969!
Gold inlaid with almandines, pearls, glass, and enamel. Rijksmuseum van Oudheden, Leiden 📷 AlexanderVanLoon www.rmo.nl/en/collectio...
The Dorestad Brooch, c. AD 800.
This beautiful Carolingian brooch was found at the bottom of a well in the Netherlands in 1969!
Gold inlaid with almandines, pearls, glass, and enamel. Rijksmuseum van Oudheden, Leiden 📷 AlexanderVanLoon www.rmo.nl/en/collectio...
I bet builders were the same in ancient Babylon too - give you a quote, do the work (late), then bill you for much more.
Having nothing now in my mind of trouble in the world, but quite the contrary, much joy, except only the ending of our difference with my uncle Thomas, and the getting of the bills well over for my building of my house here.
January 3, 2026 at 6:47 PM
I bet builders were the same in ancient Babylon too - give you a quote, do the work (late), then bill you for much more.
The Saint-Exupéry 50FF note had a security mark in the shape of a fluorescent purple elephant that had been eaten by a boa constrictor. Or maybe it was a hat. I have an unused note somewhere.
Je n'éprouve aucune nostalgie du Franc comme monnaie mais je regrette cette époque où les billets de banques étaient tout simplement beaux...
Saint-Exupéry, Delacroix, Montesquieu, Pascal...
Ça avait tout de même infiniment plus d'allure que les horribles billets en Euros qu'ils nous ont pondus.
Saint-Exupéry, Delacroix, Montesquieu, Pascal...
Ça avait tout de même infiniment plus d'allure que les horribles billets en Euros qu'ils nous ont pondus.
January 3, 2026 at 6:41 PM
The Saint-Exupéry 50FF note had a security mark in the shape of a fluorescent purple elephant that had been eaten by a boa constrictor. Or maybe it was a hat. I have an unused note somewhere.
Reposted by Matthew Cobb
January 3, 2026 at 4:16 PM
Reposted by Matthew Cobb
Come and join me and my colleagues at the Department of Biology, #LundUniversity in #sweden! We have am open position as Senior Lecturer/Associate Professor in Biodiversity.
Apply here no later than February 11 2026:
https://lu.varbi.com/en/what:job/jobID:848749/type:job/where:4/apply:1
Apply here no later than February 11 2026:
https://lu.varbi.com/en/what:job/jobID:848749/type:job/where:4/apply:1
Senior Lecturer/Associate Professor in Biodiversity
The Department of Biology was established in 2010 through the merger of the Departments of Ecology, Cell and Organism Biology, Biological Undergraduate Education, and the Biological Museums. The depar
lu.varbi.com
December 20, 2025 at 3:58 PM
Come and join me and my colleagues at the Department of Biology, #LundUniversity in #sweden! We have am open position as Senior Lecturer/Associate Professor in Biodiversity.
Apply here no later than February 11 2026:
https://lu.varbi.com/en/what:job/jobID:848749/type:job/where:4/apply:1
Apply here no later than February 11 2026:
https://lu.varbi.com/en/what:job/jobID:848749/type:job/where:4/apply:1
Reposted by Matthew Cobb
Male Skate, ctenophore, & sea pen. @schmidtocean.bsky.social dive 891. #argentiniandeepseeps #CONICET #MarineLife
January 3, 2026 at 1:09 PM
Male Skate, ctenophore, & sea pen. @schmidtocean.bsky.social dive 891. #argentiniandeepseeps #CONICET #MarineLife
Reposted by Matthew Cobb
Two images, one taken near perihelion, and again near aphelion showing the difference in apparent size (about 3% difference between them). A fixed optical system (Seestar 50) means the image scale is constant.
January 3, 2026 at 12:28 PM
Two images, one taken near perihelion, and again near aphelion showing the difference in apparent size (about 3% difference between them). A fixed optical system (Seestar 50) means the image scale is constant.
Reposted by Matthew Cobb
Reposted by Matthew Cobb
2/ ~old jk is best jk~
January 3, 2026 at 10:12 AM
2/ ~old jk is best jk~
We’re here again. Grenada 1983. Panama 1989. Nicaragua clandestinely throughout the 1980s. The US “back yard”.
January 3, 2026 at 10:22 AM
We’re here again. Grenada 1983. Panama 1989. Nicaragua clandestinely throughout the 1980s. The US “back yard”.
Hard to post any of the stuff I had in mind when this kind of thing is happening.
January 3, 2026 at 9:39 AM
Hard to post any of the stuff I had in mind when this kind of thing is happening.
Reposted by Matthew Cobb
The problem is, if you're just a lay person googling a question or asking for a picture of a cat with a bird on its head, mostly you'll see GenAI stuff that seems fine.
If you're a subject matter expert, the vast majority of GenAI products you encounter are absolute slop that wastes your time.
If you're a subject matter expert, the vast majority of GenAI products you encounter are absolute slop that wastes your time.
Microsoft CEO Satya Nadella is now blogging about AI slop. Nadella wants everyone to move beyond the arguments about AI slop, in a "pivotal year for AI". Full details here 👇 www.theverge.com/news/852630/...
Microsoft CEO Satya Nadella is now blogging about AI slop
Nadella doesn’t want to argue about AI slop anymore
www.theverge.com
January 2, 2026 at 8:47 PM
The problem is, if you're just a lay person googling a question or asking for a picture of a cat with a bird on its head, mostly you'll see GenAI stuff that seems fine.
If you're a subject matter expert, the vast majority of GenAI products you encounter are absolute slop that wastes your time.
If you're a subject matter expert, the vast majority of GenAI products you encounter are absolute slop that wastes your time.
Reposted by Matthew Cobb
HiPOD mini: Ancient Channels or Valleys
This image shows us an ancient channel, perhaps an ancient valley, on the surface of Mars. We can also see that the terrain is heavily pockmarked with impact craters. The scene is less than 5 km, or 3 miles, across.
uahirise.org/ESP_076586_1...
#Mars
This image shows us an ancient channel, perhaps an ancient valley, on the surface of Mars. We can also see that the terrain is heavily pockmarked with impact craters. The scene is less than 5 km, or 3 miles, across.
uahirise.org/ESP_076586_1...
#Mars
January 2, 2026 at 8:40 PM
HiPOD mini: Ancient Channels or Valleys
This image shows us an ancient channel, perhaps an ancient valley, on the surface of Mars. We can also see that the terrain is heavily pockmarked with impact craters. The scene is less than 5 km, or 3 miles, across.
uahirise.org/ESP_076586_1...
#Mars
This image shows us an ancient channel, perhaps an ancient valley, on the surface of Mars. We can also see that the terrain is heavily pockmarked with impact craters. The scene is less than 5 km, or 3 miles, across.
uahirise.org/ESP_076586_1...
#Mars
Reposted by Matthew Cobb
I've been tracking the spread of nonconsensual deepfakes on X for more than two years. Here's a timeline of how Musk's leadership allowed the practice to flourish from a once-underground market to a viral trend, with little recourse for victims or legal enforcement.
spitfirenews.com/p/grok-csam-...
spitfirenews.com/p/grok-csam-...
How Grok's sexual abuse hit a tipping point
Nonconsensual deepfakes on X are nothing new, but now it's built into the platform.
spitfirenews.com
January 2, 2026 at 7:47 PM
I've been tracking the spread of nonconsensual deepfakes on X for more than two years. Here's a timeline of how Musk's leadership allowed the practice to flourish from a once-underground market to a viral trend, with little recourse for victims or legal enforcement.
spitfirenews.com/p/grok-csam-...
spitfirenews.com/p/grok-csam-...
Reposted by Matthew Cobb
Parker's piece nails why this matters. "Tech companies... would rather not answer for their products’ failures. Every headline that says “Grok apologizes” or “Grok admits” or “Grok says” creates a world where the chatbot takes the fall while Musk and his executives face no scrutiny whatsoever."
I wrote about Grok's CSAM problem and how journalists keep fumbling the story
www.readtpa.com/p/grok-cant-...
www.readtpa.com/p/grok-cant-...
Grok Can't Apologize. Grok Isn't Sentient. So Why Do Headlines Keep Saying It Did?
Elon Musk’s chatbot has been “undressing” women and children in public. Journalists are letting xAI off the hook by pretending the bot is responsible.
www.readtpa.com
January 2, 2026 at 7:32 PM
Parker's piece nails why this matters. "Tech companies... would rather not answer for their products’ failures. Every headline that says “Grok apologizes” or “Grok admits” or “Grok says” creates a world where the chatbot takes the fall while Musk and his executives face no scrutiny whatsoever."
Reposted by Matthew Cobb
there are endless headlines about "grok apologizes" or "grok says it will be fixed" after it generated CSAM using pictures of minors and zero headlines about "grok being shut down" or "musk and xAI in big legal trouble" and I think something fundamental has broken here
January 2, 2026 at 6:59 PM
there are endless headlines about "grok apologizes" or "grok says it will be fixed" after it generated CSAM using pictures of minors and zero headlines about "grok being shut down" or "musk and xAI in big legal trouble" and I think something fundamental has broken here
Reposted by Matthew Cobb
quoting chatbots as official company sources shows that our tech reporters are not up to the task of understanding the basics of the technology they're reporting on
January 2, 2026 at 6:35 PM
quoting chatbots as official company sources shows that our tech reporters are not up to the task of understanding the basics of the technology they're reporting on
Reposted by Matthew Cobb
Giant moray
#gili #giliislands #lombok #diving #scuba #trawangan #diveandstay #giliair #pets #ocean #sealife #marinelife #padi #morayeel #eel #moray
#gili #giliislands #lombok #diving #scuba #trawangan #diveandstay #giliair #pets #ocean #sealife #marinelife #padi #morayeel #eel #moray
January 2, 2026 at 7:03 PM
Reposted by Matthew Cobb
I never caught Chickenpox as a kid and was exposed to shingles when I was pregnant. Had to have antivirals to protect my baby from serious harm or death. Finally, I caught it when I was 37 years old, and I've never been so sick or experienced pain like it. Please vaccinate your children.
Woohoo! Chickenpox vaccines are on the menu in the UK - finally!!
www.theguardian.com/commentisfre...
www.theguardian.com/commentisfre...
Ignore the sceptics: with this new vaccine, chickenpox could become a thing of the past | Wes Streeting
Thousands develop serious complications from the varicella virus. Today the NHS rolls out a new jab, alongside measles, mumps and rubella, says health secretary Wes Streeting
www.theguardian.com
January 2, 2026 at 2:37 PM
I never caught Chickenpox as a kid and was exposed to shingles when I was pregnant. Had to have antivirals to protect my baby from serious harm or death. Finally, I caught it when I was 37 years old, and I've never been so sick or experienced pain like it. Please vaccinate your children.
Reposted by Matthew Cobb
Several French government ministers have reported the unlawful production of sexualised material without consent to the police and prosecutors. The UK government, which published a high profile strategy on misogyny just before Christmas, has yet to say or do anything
www.politico.eu/article/fran...
www.politico.eu/article/fran...
France to investigate deepfakes of women stripped naked by Grok
Hundreds of women and teenagers have reported their photos published on social media have been “undressed” by Grok.
www.politico.eu
January 2, 2026 at 5:46 PM
Several French government ministers have reported the unlawful production of sexualised material without consent to the police and prosecutors. The UK government, which published a high profile strategy on misogyny just before Christmas, has yet to say or do anything
www.politico.eu/article/fran...
www.politico.eu/article/fran...
Reposted by Matthew Cobb
AI cannot take accountability. AI cannot be “sorry.”
You have no control over the outputs of AI if you did not make it yourself, train it on very specific data sets, and create extremely rigorous control methods. Even then the safeguards are limited.
The creators must be held accountable
You have no control over the outputs of AI if you did not make it yourself, train it on very specific data sets, and create extremely rigorous control methods. Even then the safeguards are limited.
The creators must be held accountable
January 2, 2026 at 5:23 PM
AI cannot take accountability. AI cannot be “sorry.”
You have no control over the outputs of AI if you did not make it yourself, train it on very specific data sets, and create extremely rigorous control methods. Even then the safeguards are limited.
The creators must be held accountable
You have no control over the outputs of AI if you did not make it yourself, train it on very specific data sets, and create extremely rigorous control methods. Even then the safeguards are limited.
The creators must be held accountable