Is ChatGPT making scientists hyper-productive? The highs and lows of using AI

"Large language models are transforming scientific writing and publishing. But the productivity boost that these tools bring could have a downside.

 

ChatGPT continues to steal the spotlight, more than a year after its public debut.

 

The artificial intelligence (AI) chatbot was released as a free-to-use tool in November 2022 by technology company OpenAI in San Francisco, California. Two months later, ChatGPT had already been listed as an author on a handful of research papers.

 

ChatGPT one year on: who in academy is using it, how and why?

 

Academic publishers scrambled to announce policies on the use of ChatGPT and other large language models (LLMs) in the writing process. By last October, 87 out of 100 top scientific journals had provided guidance to authors on the use of generative AI, which can create text, images and other content, researchers reported on 31 January in the The BMJ1.

 

But that’s not the only way in which ChatGPT and other LLMs have begun to change scientific writing. In academia’s competitive environment, any tool that allows researchers to “produce more publications is going to be a very attractive proposition”, says digital-innovation researcher Savvas Papagiannidis at Newcastle University in Newcastle upon Tyne, UK.

 

Generative AI is continuing to improve — so publishers, grant-funding agencies and scientists must consider what constitutes ethical use of LLMs, and what over-reliance on these tools says about a research landscape that encourages hyper-productivity.

Are scientists routinely using LLMs to write papers?

 

Before its public release, ChatGPT was not nearly as user-friendly as it is today, says computer scientist Debora Weber-Wulff at the HTW Berlin University of Applied Sciences. “The interfaces for the older GPT models were something that only a computer scientist could love.”

 

In the past, researchers typically needed expertise to use advanced LLMs. Now, “GPT has democratized that to some degree”, says Papagiannidis.

 

This democratization has catalysed the use of LLMs in research writing. In a 2023 Nature survey of more than 1,600 scientists, almost 30% said that they had used generative AI tools to help write manuscripts, and about 15% said they had used them to help write grant applications.

 

And LLMs have many other uses. They can help scientists to write code, brainstorm research ideas and conduct literature reviews. LLMs from other developers are improving, as well — Google’s Gemini, for example, and Claude 2 by Anthropic, an AI company in San Francisco.

 

 Researchers with the right skills can even develop their own personalized LLMs that are fine-tuned to their writing style and scientific field, says Thomas Lancaster, a computer scientist at Imperial College London.

 

What are the benefits for researchers?

 

About 55% of the respondents to the Nature survey felt that a major benefit of generative AI is its ability to edit and translate writing for researchers whose first language is not English. Similarly, in a poll by the European Research Council (ERC), which funds research in the European Union, 75% of more than 1,000 ERC grant recipients felt that generative AI would reduce language barriers in research by 2030, according to a report released in December2.

 

Of the ERC survey respondents, 85% thought that generative AI could take on repetitive or labour-intensive tasks, such as literature reviews. And 38% felt that generative AI would promote productivity in science, such as by helping researchers to write papers at a faster pace.

What are the downsides?

 

Although ChatGPT’s output can be convincingly human-like, Weber-Wulff warns that LLMs can still make language mistakes that readers might notice. That’s one of the reasons she advocates for researchers to acknowledge LLM use in their papers. Chatbots are also notorious for generating fabricated information, called hallucinations.

 

How ChatGPT and other AI tools could disrupt scientific publishing

 

And there is a drawback to the productivity boost that LLMs might bring. Speeding up the paper-writing process could increase throughput at journals, potentially stretching editors and peer reviewers even thinner than they already are. “With this ever-increasing number of papers — because the numbers are going up every year — there just aren’t enough people available to continue to do free peer review for publishers,” Lancaster says. He also points out that alongside researchers who openly use LLMs and acknowledge it, some quietly use the tools to churn out low-value research.

 

It’s already difficult to sift through the sea of published papers to find meaningful research, Papagiannidis says. If ChatGPT and other LLMs increase output, this will prove even more challenging.

 

“We have to go back and look at what the reward system is in academia,” Weber-Wulff says. The current ‘publish or perish’ model rewards researchers for constantly pushing out papers. But many people argue that this needs to shift towards a system that prioritizes quality over quantity. For example, Weber-Wulff says, the German Research Foundation allows grant applicants to include only ten publications in a proposal. “You want to focus your work on getting really good, high-level papers,” she says.

Where do scientific publishers stand on LLM use?

 

According to the study in The BMJ, 24 of the world’s 100 largest publishers — collectively responsible for more than 28,000 journals — had by last October provided guidance on generative AI1. Journals with generative-AI policies tend to allow some use of ChatGPT and other LLMs, as long as they’re properly acknowledged.

 

Springer Nature, for example, states that LLM use should be documented in the methods or another section of the manuscript, a guideline introduced in January 2023. Generative AI tools do not, however, satisfy criteria for authorship, because that “carries with it accountability for the work, and AI tools cannot take such responsibility”. (Nature’s news team is editorially independent of its publisher, Springer Nature.)

 

Enforcing these rules is easier said than done, because undisclosed AI-generated text can be difficult for publishers and peer reviewers to spot. Some sleuths have caught it through subtle phrases and mistranslations. 

 

Unlike cases of plagiarism, in which there is clear source material, “you can’t prove that anything was written by AI”, Weber-Wulff says. 

 

Despite researchers racing to create LLM-detection tools, “we haven’t seen one that we thought produced a compelling enough result” to screen journal submissions, says Holden Thorp, editor-in-chief of the Science family of journals.

What about other uses?

 

Although as of November, the American Association for the Advancement of Science — which publishes Science — allows for some disclosed use of generative AI in the preparation of manuscripts, it still bans the use of LLMs during peer review, Thorp says. This is because he and others at Science want reviewers to devote their full attention to the manuscript being assessed, he adds. Similarly, Springer Nature’s policy prohibits peer reviewers from uploading manuscripts into generative-AI tools.

 

Scientific sleuths spot dishonest ChatGPT use in papers

 

Some grant-funding agencies, including the US National Institutes of Health and the Australian Research Council, forbid reviewers from using generative AI to help examine grant applications because of concerns about confidentiality (grant proposals are treated as confidential documents, and data entered into public LLMs could be accessed by other people).

 

 But the ERC Scientific Council, which governs the ERC, released a statement in December recognizing that researchers use AI technologies, along with other forms of external help, to prepare grant proposals. It said that, in these cases, authors must still take full responsibility for their work.

 

“Many organizations come out now with very defensive statements” requiring authors to acknowledge all use of generative AI, says ERC Scientific Council member Tom Henzinger, a computer scientist at the Institute of Science and Technology Austria in Klosterneuburg.

 

To him, the use of ChatGPT seems no different from running text by a colleague for feedback. “Use every resource at your disposal,” Henzinger says.

 

Regardless of the ever-changing rules surrounding generative AI, researchers will continue to use it, Lancaster says. “There is no way of policing the use of technology like ChatGPT.”" [1]

 

1. Nature 627, 16-17 (2024)   By McKenzie Prillaman

„Epigenetinis“ redagavimas sumažina pelių cholesterolio kiekį

 

 „Pelių DNR cheminių žymenų pokyčiai sumažina geno aktyvumą be genomo pjūvių.

 

 

 

 Remiantis tyrimu su pelėmis, alternatyva genomo redagavimui gali sumažinti geno, turinčio įtakos cholesterolio kiekiui, aktyvumą, nekeičiant DNR sekos – ir tai daroma ilgą laiką.

 

 

 

 Mokslininkai šį efektą pasiekė, pakeitę kiekvieno gyvūno „epigenomą“, kurio vienas bruožas yra cheminių žymenų, susietų su DNR ir turinčių įtakos genų veiklai, rinkinys. Po gydymo tikslinio geno aktyvumas sumažėjo ir išliko žemas 11 mėnesių, per kuriuos buvo tiriamos pelės.

 

 

 

 2023 m. patvirtinus pirmąją genomo redagavimo terapiją, kuri remiasi CRISPR-Cas9 redagavimo sistema, buvo sukurta nauja medicinos forma, apimanti tikslinius DNR sekų pakeitimus. Tačiau naujos išvados, paskelbtos vasario 28 d. žurnale Nature, patvirtina, kad reikia redaguoti epigenomą tam tikroms ligoms gydyti, taip išvengiant tam tikros rizikos, kylančios dėl DNR grandinių laužymo ir negrįžtamo pakeitimo.

 

 

 

 „Tai tik eros, kai reikia atsikratyti manipuliacijų su DNR, pradžia“, – sako Kalifornijos universiteto Deiviso epigenetikė Henriette O’Geen. "Tai gali pakeisti genų, susijusių su liga, ekspresiją ir, galbūt, išgydyti, nekeičiant DNR."

 

 Pažymėkite šį geną

 

 

 

 Vystymosi metu ląstelės įgyja naują tapatybę, cheminių žymenų modelis jų DNR dažnai keičiasi. Dėl šių epigenetinių pakitimų ląstelė gali elgtis kaip, pavyzdžiui, kepenų ląstelė, o ne smegenų ląstelė.

 

 

 

 Ar šis vienkartinis „epigenetinis“ gydymas galėtų kontroliuoti cholesterolio kiekį?

 

 

 

 Po daugiau, nei dešimtmetį, trukusių, pastangų mokslininkai sugalvojo, kaip modifikuoti genomo redagavimo įrankius, kad būtų pakoreguoti kai kurie epigenetiniai ženklai. Tai leidžia prie DNR tiksliose vietose pridėti cheminės žymos, vadinamos metilo grupe, tipą, pavyzdžiui, norint išjungti geną arba pašalinti metilo grupes iš genomo vietos, kad genas būtų įjungtas.

 

 

 

 Epigenetinio redagavimo taikymas klinikoje iš pradžių buvo neaiškus, sako epigenetikė Marianne Rots iš Groningeno universiteto medicinos centro Nyderlanduose. Tyrėjai buvo susirūpinę dėl to, koks konkretus ar veiksmingas bus šis metodas, ir kiek ilgai truks jo poveikis.

 

 Pirštas ant genomo

 

 

 

 Norėdami išspręsti šiuos klausimus, Angelo Lombardo, Milano (Italija) mokslinio instituto San Raffaele genų terapijos tyrinėtojas, ir jo kolegos naudojo molekules, vadinamas cinko piršto baltymais, kurios, panašiai kaip CRISPR-Cas9 sistema, gali būti suprojektuotos taip, kad prisijungtų prie specifinės sekos genome. Komanda sukūrė cinko piršto baltymą, kuris galėtų prisijungti prie PCSK9 geno, kuris yra kelių esamų aukšto cholesterolio kiekio gydymo būdų taikinys. Tada autoriai suliejo savo cinko pirštų baltymus su trijų baltymų, susijusių su metilo grupių prijungimu prie DNR, gabalėliais.

 

 

 

 Tas fragmentų kokteilis buvo paimtas iš baltymų, veikiančių embriono vystymosi metu, rinkinio, pridedant metilo grupių, siekiant užtikrinti, kad genome slypinčios virusų sekos – praeities infekcijų reliktai – būtų nutildytos ir tokios išliktų visą gyvenimą. Lombardo teigimu, tikimasi, kad ilgalaikis šio natūralaus epigenetinio redagavimo poveikis bus perkeltas į geną, susietą su cinko piršto baltymu, kurį sukūrė autoriai.

 

 

 

 Dirbdama su pelėmis, komanda naudojo šią sistemą Pcsk9 genui redaguoti. Gyvūnų cholesterolio kiekis sumažėjo per mėnesį po gydymo. Jų PCSK9 baltymo lygis taip pat sumažėjo ir išliko žemas 330 dienų, kurias tyrėjai juos stebėjo. Poveikis gali trukti ilgiau, nei metus, sako O'Geenas, atsižvelgiant į tai, kad eksperimento pabaigoje graužikų PSCK9 lygis nerodė jokių atsigavimo ženklų.

 

 Skubėjimas į epigenetiką

 

 

 

 Rezultatai papildys jau augantį jaudulį dėl epigenetinio redagavimo. Daugiau, nei dešimt kompanijų yra orientuotos į epigenetinio redagavimo terapijų kūrimą, sako Rots. Kai kurios pranešė apie ilgalaikį poveikį beždžionėms, tačiau dar nepaskelbė savo išvadų recenzuojamuose žurnaluose.

 

 

 

 Omega Therapeutics, Kembridže, Masačusetso valstijoje, atlieka klinikinį epigenetinio redaktoriaus tyrimą, kuris nutildo MYC – geną, kuris yra pernelyg aktyvus, sergant daugeliu vėžio formų ir kurį sunku pasiekti, naudojant įprastus vaistus. „Įdomu matyti, kaip viskas sprogo“, – sako Rotsas." [1]

 

1. Nature 627, 14-15 (2024) By Heidi Ledford

‘Epigenetic’ editing cuts cholesterol in mice

 

"Changes to chemical tags on DNA in mice dial down the activity of a gene without cuts to the genome.

 

An alternative to genome editing can reduce the activity of a gene that affects cholesterol levels without changing the DNA sequence — and does so for an extended period, according to a study in mice.

 

Scientists achieved this effect by changing each animal’s ‘epigenome’, one feature of which is a collection of chemical tags that are bound to DNA and affect gene activity. After the treatment, activity of the targeted gene dropped and remained low for the 11 months over which the mice were studied.

 

The 2023 approvals of the first genome-editing therapy, which relies on the CRISPR–Cas9 editing system, ushered in a new form of medicine that involves making targeted changes to DNA sequences. But the new findings, published on 28 February in Nature, bolster the case for instead editing the epigenome to treat certain diseases, thereby sidestepping some of the risks that come with breaking and irreversibly altering strands of DNA.

 

“This is just the beginning of an era of getting away from cutting DNA,” says Henriette O’Geen, an epigeneticist at the University of California, Davis. “This can alter the expression of genes that are involved in disease — and potentially provide a cure — without changing DNA.”

Mark this gene

 

As cells take on new identities during development, the pattern of chemical tags on their DNA often changes. These epigenetic alterations can tell a cell to behave as a liver cell, for example, rather than a brain cell.

 

Could this one-time ‘epigenetic’ treatment control cholesterol?

 

After more than a decade of effort, scientists worked out how to modify genome-editing tools to tweak some epigenetic marks. This makes it possible to add a type of chemical tag called a methyl group to DNA at precise locations, for example, to switch a gene off, or to remove methyl groups from a spot in the genome to turn a gene on2.

 

Epigenetic editing’s applications in the clinic were initially unclear, says epigeneticist Marianne Rots at the University Medical Center Groningen in the Netherlands. Researchers were concerned about how specific or effective the approach would be, she says, and how long its effects would last.

A finger on the genome

 

To address these questions, Angelo Lombardo, a gene-therapy researcher at the San Raffaele Scientific Institute in Milan, Italy, and his colleagues used molecules called zinc-finger proteins that, much like the CRISPR–Cas9 system, can be designed to bind to specific sequences in the genome. The team designed a zinc-finger protein that could bind to the PCSK9 gene, which is the target of several existing therapies for high cholesterol. The authors then fused their zinc-finger proteins with pieces of three proteins involved in attaching methyl groups to DNA.

 

That cocktail of fragments was drawn from a suite of proteins that act during embryonic development, adding methyl groups to ensure that viral sequences lurking in the genome — relics of past infections — are silenced and stay that way for a lifetime. The hope, says Lombardo, was that the long-lasting effects of that natural epigenetic editing would carry over to the gene bound by the zinc-finger protein that the authors designed.

 

Working with mice, the team used this system to edit the Pcsk9 gene. The animals’ cholesterol levels fell within a month of the treatment. Their levels of the PCSK9 protein also dropped — and stayed low for the 330 days that the researchers tracked them. The effects could last longer than a year, says O’Geen, given that the rodents’ PSCK9 levels showed no signs of rebounding at the end of the experiment.

A rush to epigenetics

 

The results will add to already burgeoning excitement about epigenetic editing. More than ten companies are focused on developing epigenetic editing therapies, says Rots. A few have reported long-lasting effects in monkeys but have not yet published their findings in peer-reviewed journals.

 

And Omega Therapeutics, a company in Cambridge, Massachusetts, is conducting a clinical trial of an epigenetic editor that silences MYC, a gene that is overactive in many cancers and has been difficult to target using conventional drugs. “It’s exciting to see how things have exploded,” Rots says." [1]

 

1. Nature 627, 14-15 (2024) By Heidi Ledford

 

Taxes paid by Lithuanian businesses are too low.

 

Much smaller than other Western European countries. This is not a plus, it is a problem. Reviving Lithuania is difficult, it requires money. We can no longer compete with cheap shots, and we do not prepare good shots. The health care system is in a sorry state. Most of the pensioners are starving (kefir and potatoes). Law enforcement does not work: the Landsbergs and their henchmen openly steal and scoff. State management is catastrophic - we are the only ones in the Western world who have declared war and territorial claims against China, we, Lithuanians, claim Taiwan, territory of China.

 Vidmantas Janulevičius, the president of the Confederation of Lithuanian Industrialists (LPK), proposes that we do not change taxes in 2024-2026 . Then nothing would change. So how could it work now?

 

 

Lietuvos verslo mokami mokesčiai yra pernelyg maži.

 

Daug mažesni, nei kitų Vakarų Europos šalių. Tai nėra pliusas, tai yra problema. Lietuvą prikelti sunku, tam reikia pinigų. Pigiais kadrais konkuruoti jau negalime, o gerų kadrų neparuošiame. Sveikatos apsaugos sistema yra apgailėtinos būklės. Pensininkų dauguma badauja (kefyras ir bulvės). Teisėsauga neveikia: Landsbergiai ir jų pakalikai atvirai vagia ir šaiposi. Valstybės valdymas katastrofiškas - mes vieninteliai Vakarų pasaulyje paskelbėme karą ir teritorines pretenzijas Kinijai, pretenduojame į jos Taivaną.

 Vidmantas Janulevičius, Lietuvos pramonininkų konfederacijos (LPK) prezidentas, siūlo 2024–2026 m. nekeisti mokesčių. Tada niekas nesikeistų. Tai kaip čia dabar išeina?