Samuli Reijula
@samulireijula.fediscience.org.ap.brid.gy
PhilSci & Cogsci at University of Helsinki / TINT. Carbon-based stochastic parrot? European Network for the Science of Science (www.euroscisci.net) […]
[bridged from https://fediscience.org/@samulireijula on the fediverse by https://fed.brid.gy/ ]
[bridged from https://fediscience.org/@samulireijula on the fediverse by https://fed.brid.gy/ ]
Reposted by Samuli Reijula
As someone here recently pointed out, this will not be a long-term problem. In a couple of years, fake citations will be easy to spot. So as a researcher, you want to be careful now, to protect your long-term reputation. So for now, don’t rely on shady sources. +
AI causes so much extra work. You have to check and double check everything.
I’m sorry, but it is disgraceful to be an academic who uses this technology to conduct research. It should be prohibited in all of our scholarly institutions, including universities and journals.
December 21, 2025 at 8:36 AM
As someone here recently pointed out, this will not be a long-term problem. In a couple of years, fake citations will be easy to spot. So as a researcher, you want to be careful now, to protect your long-term reputation. So for now, don’t rely on shady sources. +
Reposted by Samuli Reijula
+ "Pitäisikö nyt keskustella tiedon sääntelystä? - ”Olisin ehdottomasti sen kannalla”, Riie Heikkilä sanoo. ” Puhutaanko tässä epistemologiasta? Vai sääntelystä jotenkin konkreettisemmassa yhteiskunnallisessa mielessä, siitä kuka saa esittää tietoväitteitä?
December 30, 2025 at 9:49 AM
+ "Pitäisikö nyt keskustella tiedon sääntelystä? - ”Olisin ehdottomasti sen kannalla”, Riie Heikkilä sanoo. ” Puhutaanko tässä epistemologiasta? Vai sääntelystä jotenkin konkreettisemmassa yhteiskunnallisessa mielessä, siitä kuka saa esittää tietoväitteitä?
Reposted by Samuli Reijula
tästä tärkeästä jutusta on puhuttu monta päivää, mutta jatkan. mitä hittoa on "tiedon sääntely"? www.hs.fi/kulttuuri/ar... +
Essee | ”Sivistynyt työväenluokka on peruuttamattomasti loppu”, sanoo tutkija
Tieto on nyt arvokasta, ja siksi sitä eivät saa kaikki. Tunne epätasa-arvosta on ollut historiallisesti yleisin syy kumouksille ja konflikteille, kirjoittaa Jussi Ahlroth.
www.hs.fi
December 30, 2025 at 9:49 AM
tästä tärkeästä jutusta on puhuttu monta päivää, mutta jatkan. mitä hittoa on "tiedon sääntely"? www.hs.fi/kulttuuri/ar... +
Reposted by Samuli Reijula
Pakolaisleirien pommittaminen ei missään olosuhteissa voi olla hyväksyttävää. Näin ulkopuolelta taviksena ei tietenkään ymmärtää mitä kaikkea on tekeillä mutta Netanjahu pitää saada pois vallasta
October 14, 2024 at 5:39 PM
Pakolaisleirien pommittaminen ei missään olosuhteissa voi olla hyväksyttävää. Näin ulkopuolelta taviksena ei tietenkään ymmärtää mitä kaikkea on tekeillä mutta Netanjahu pitää saada pois vallasta
Reposted by Samuli Reijula
Next Monday, all welcome! #philsci
In the next Perspectives on Science seminar, Benjamin Santos Genta (University of California, Irvine) will give a talk titled “The Easy and Hard Problem of Similarity”. Find more details and join the seminar via tint-helsinki.fi?p=1354
7.10. Perspectives on Science seminar: Benjamin Santos Genta – Centre for Philosophy of Social Science
tint-helsinki.fi
October 4, 2024 at 11:19 AM
Next Monday, all welcome! #philsci
Reposted by Samuli Reijula
tekoäly on usein hieno juttu mutta kilpajuoksu deepfakejen tunnistamisen kanssa on jo hävitty. sen sijaan self-authenticating documents ja sisältöjen allekirjoittaminen lupaava juttu www.law.cornell.edu/rules/fre/ru... @jussipullinen.bsky.social www.hs.fi/paakirjoituk...
Kolumni | Suomalaisia uhkaa digiläävä, jossa meitä huijataan
Meillä on vaikeuksia ottaa verkkohuijaukset vakavasti, vaikka ne lisääntyvät vuosi vuodelta ja tekoäly lisää niiden riskejä.
www.hs.fi
October 6, 2024 at 5:49 AM
tekoäly on usein hieno juttu mutta kilpajuoksu deepfakejen tunnistamisen kanssa on jo hävitty. sen sijaan self-authenticating documents ja sisältöjen allekirjoittaminen lupaava juttu www.law.cornell.edu/rules/fre/ru... @jussipullinen.bsky.social www.hs.fi/paakirjoituk...
Reposted by Samuli Reijula
ok i read this knightcolumbia.org/content/prot... and liked it a lot. what should I read next? #fediverse
Protocols, Not Platforms: A Technological Approach to Free Speech
knightcolumbia.org
September 30, 2024 at 6:05 AM
ok i read this knightcolumbia.org/content/prot... and liked it a lot. what should I read next? #fediverse
Reposted by Samuli Reijula
philosophers in a bar under the sea. after Kristina fest. great talks, and a festschrift was announced, an edited volume on postcolonialist and feminist philosophy of science, an area where Kristina (Rolin) has made significant contributions #philsci
September 27, 2024 at 5:45 AM
philosophers in a bar under the sea. after Kristina fest. great talks, and a festschrift was announced, an edited volume on postcolonialist and feminist philosophy of science, an area where Kristina (Rolin) has made significant contributions #philsci
Reposted by Samuli Reijula
Because the weights would be integrated into the hardware, it would be impossible to copy them from one machine to another; each AI would have to learn on its own. “They would have to go to school,” ... “But you would go from using a megawatt to 30 watts.” www.newyorker.com/magazine/202...
Why the Godfather of A.I. Fears What He’s Built
Geoffrey Hinton has spent a lifetime teaching computers to learn. Now he worries that artificial brains are better than ours.
www.newyorker.com
September 29, 2024 at 5:54 AM
Because the weights would be integrated into the hardware, it would be impossible to copy them from one machine to another; each AI would have to learn on its own. “They would have to go to school,” ... “But you would go from using a megawatt to 30 watts.” www.newyorker.com/magazine/202...
Reposted by Samuli Reijula
a beautiful idea from Hinton: "In [the brain], the weights are built into the physical connections between neurons. Hinton worked to create an artificial version of this system using specialized computer chips. [...] The chips would be able to learn by varying their “conductances.”
September 29, 2024 at 5:52 AM
a beautiful idea from Hinton: "In [the brain], the weights are built into the physical connections between neurons. Hinton worked to create an artificial version of this system using specialized computer chips. [...] The chips would be able to learn by varying their “conductances.”
Reposted by Samuli Reijula
Different studies attempting to answer the same question don't always reach the same conclusion. Sometimes it's error, sometimes noise, but on some occasion, the discordance between the findings persists. How can that be? www.journals.uchicago.edu/doi/10.1086/... #philsci #metasci
Persistent evidential discordance | The British Journal for the Philosophy of Science: Vol 0, No ja
www.journals.uchicago.edu
August 5, 2024 at 7:28 AM
Different studies attempting to answer the same question don't always reach the same conclusion. Sometimes it's error, sometimes noise, but on some occasion, the discordance between the findings persists. How can that be? www.journals.uchicago.edu/doi/10.1086/... #philsci #metasci
Reposted by Samuli Reijula
Are metascience and science of science the same thing? In my opinion, not exactly. There wasn't a feed for the science of science here, so I set one up. It picks up posts with 'science of science', #scisci, but also #metascience since much of the discussion is relevant to scisci
August 22, 2024 at 8:22 AM
Are metascience and science of science the same thing? In my opinion, not exactly. There wasn't a feed for the science of science here, so I set one up. It picks up posts with 'science of science', #scisci, but also #metascience since much of the discussion is relevant to scisci
Reposted by Samuli Reijula
another interesting piece today, citation manipulation: www.nature.com/articles/d41... #metascience
The citation black market: schemes selling fake references alarm scientists
The ways in which researchers can artificially inflate their reference counts are growing. The ways in which researchers can artificially inflate their reference counts are growing.
www.nature.com
August 21, 2024 at 11:02 AM
another interesting piece today, citation manipulation: www.nature.com/articles/d41... #metascience
Reposted by Samuli Reijula
"[R]esearchers in disciplines such as experimental psychology, microbiology and evolutionary biology experienced the most egalitarian citation distribution, while those in philosophy, education and anthropology the most unequal." #philsky #metascience royalsocietypublishing.org/doi/10.1098/...
August 21, 2024 at 6:02 AM
"[R]esearchers in disciplines such as experimental psychology, microbiology and evolutionary biology experienced the most egalitarian citation distribution, while those in philosophy, education and anthropology the most unequal." #philsky #metascience royalsocietypublishing.org/doi/10.1098/...
Reposted by Samuli Reijula
“Evidence is not a 2-place relation ‘e is evidence for h’. Rather,
Evidence is a 3-place relation.£
“The third place is to be filled by a good argument, a view, called ‘the argument theory of evidence’…This theory maintains that e is evidence for h relative to a good argument from e to h.”
Evidence is a 3-place relation.£
“The third place is to be filled by a good argument, a view, called ‘the argument theory of evidence’…This theory maintains that e is evidence for h relative to a good argument from e to h.”
August 4, 2024 at 11:40 AM
“Evidence is not a 2-place relation ‘e is evidence for h’. Rather,
Evidence is a 3-place relation.£
“The third place is to be filled by a good argument, a view, called ‘the argument theory of evidence’…This theory maintains that e is evidence for h relative to a good argument from e to h.”
Evidence is a 3-place relation.£
“The third place is to be filled by a good argument, a view, called ‘the argument theory of evidence’…This theory maintains that e is evidence for h relative to a good argument from e to h.”
not surprising but not cool at all. publishing is technically so easy now that we (academia) should really take it in our own hands https://www.thebookseller.com/news/academic-authors-shocked-after-taylor--francis-sells-access-to-their-research-to-microsoft-ai
July 24, 2024 at 8:39 AM
not surprising but not cool at all. publishing is technically so easy now that we (academia) should really take it in our own hands https://www.thebookseller.com/news/academic-authors-shocked-after-taylor--francis-sells-access-to-their-research-to-microsoft-ai
Happy to set my laptop on this desk, MPIB it’s been a while!
July 22, 2024 at 10:56 AM
Happy to set my laptop on this desk, MPIB it’s been a while!
The failed migration of academic twitter https://arxiv.org/abs/2406.04005 thoughts: as colleagues studying social media pointed out, the time of one big platform is in the past. But it’s not only that there several disconnected platforms - it seems people (=academics) simply lost interest and […]
Original post on fediscience.org
fediscience.org
June 28, 2024 at 10:54 AM
The failed migration of academic twitter https://arxiv.org/abs/2406.04005 thoughts: as colleagues studying social media pointed out, the time of one big platform is in the past. But it’s not only that there several disconnected platforms - it seems people (=academics) simply lost interest and […]