Academic Humanizer：让 AI 写的论文更像人写的 Claude 技能

ai and ml

Startup insists it's not trying to help anyone cheat the system - honest!

It's bad enough that you used AI to write a research paper instead of composing it yourself. Now, you can take the extra step to hide the evidence of your sloth. A startup has decided academics need a way to hide LLM tells, yet they insist their goal isn’t to support bad habits among boffins.

AI humanizers are nothing new - take a cursory look online and you’ll find that companies pushing AI that helps AI writing sound less like AI wrote it are a dime a dozen. None, says the team behind Academic Humanizer, have been specifically tailored to papers and grant proposals. Thankfully for the world’s lazy academics unwilling to actually do their jobs, that’s now changed.

“The problem is that AI-assisted drafts come out generic and verbose, with … inflated phrasing, and over-long sentences,” University of Minnesota associate professor and cofounder of MorphMind, the company behind the tool, Jie Ding, wrote in its GitHub readme. “They also drift from the author's own voice and lose the precision scholarship depends on.”

A previous version of that statement, written prior to a recent readme update, lays the point of the tool bare. Academic Humanizer, the opening of the readme read until an update five days ago, is designed to “strip the tells of AI writing from papers and grant proposals, without flattening the precise, evidence-bound voice scholarship requires.”

The GitHub repo includes several examples of text run through the humanizer for those who want to see what it’s actually capable of.

Academic Humanizer is, at its core, just a Claude skill, so it’s using more AI to correct the problems introduced into the academic writing space by AI. To further make its generative utterances more lifelike, Academic Humanizer can be pointed directly at its user’s prior work so that it’ll sound more like the “author” of the new paper instead of a bot.

As the GitHub repo and Ding both point out, the Humanizer is designed to rewrite AI-assisted drafts rather than generate new content, so it isn't supposed to pile more AI slop onto an already AI-generated draft.

“Academic Humanizer is a writing-clarity tool to help researchers express their own ideas more precisely. It is not designed to generate novel content or circumvent review,” Ding told The Register in an email. The readme further warns users that Academic Humanizer doesn’t remove their obligation to disclose that AI helped them write a paper, or wrote it in its entirety.

But if the tool isn't capable of generating findings, inventing data, or changing citations, instead serving only as "an editing aid for clarity and voice," what does that mean for weak or error-prone material in the original AI-generated draft? It probably means that while the tool may tighten prose, it isn't intended to verify the underlying arguments or evidence. Weak content may simply end up sounding more convincingly human - not a great prospect for an industry already drowning in AI-generated slop.

Academics at The University of Surrey warned last year that they were being inundated by “formulaic research articles” with “superficial and oversimplified … analysis” that were obviously crafted by large language models. The problem has even reached the AI research space itself, where last month AI output detector GPTZero said it found 100 hallucinated references across 51 papers accepted by the Conference on Neural Information Processing Systems (NeurIPS).

MORE CONTEXT

AI gets more 'meh' as you get to know it better, researchers discover; Scholars sneaking phrases into papers to fool AI reviewers; AI spies questionable science journals, with some human help; Some scientists can't stop using AI to write research papers

It doesn’t bode well for academic integrity, already imperiled by the growing use of AI in higher education and academia, nor does it bode well for the quality of college graduates, either. According to MIT, students who use AI to write essays and papers have less brain activity than those who forego the tools, leading to poorer retention and less actual learning.

But sure, let’s create a tool to make AI writing sound more human - that’s definitely the right approach to take in this situation. ®

Fire up that iPhone 1 emulation and party like it's 2007

Browser-based fun to remember a time before Apple's annual cash grab became a thing

New motherboards, old RAM, plenty of life left in yesterday's PCs

DDR4 makes a comeback as component costs rise, while Linux offers an escape from forced upgrades

The VMware exit is a protection upgrade

PARTNER CONTENT: VergeIO says an exit is also an opportunity

Tiny11 puts Windows 11 on a diet without the PowerShell workout

New desktop app simplifies stripping out bundled software, with heavier cuts carrying security and support risks

It won't kill you, or humanity, to get off the AI upgrade treadmill

New models arrive so fast, we already have more new tech than it's possible to understand in a career

Cisco, Nvidia team up to build fatter, faster pluggable

If you thought 1.6 Tbps was fast, just wait, Cisco's USFP targets 12.8 Tbps in a single optical pluggable

TOP STORIES

Anthropic asks users to stop being mean to Claude; '123456' password linked to Denmark data breach affecting 8.8M people; Microsoft and Anthropic play invoice tennis with startup's $17,600 Claude bill; Starlink's plan to avoid orbital near-misses: Ephemeris sharing; Shai-Hulud worm makes jump to AI infrastructure with Tensorlake compromise; Dutch tax office ditches Microsoft 365 cloud for on-premises alternative

Cisco, Nvidia team up to build fatter, faster pluggable If you thought 1.6 Tbps was fast, just wait, Cisco's USFP targets 12.8 Tbps in a single optical pluggable; Claude submitted fake murder tip to Philly police during summertime testing Anthropic took more than two months to alert cops; AI systems are fully capable of carrying out nightmare attacks against infrastructure and nobody's ready Meanwhile, AI agents move robotic arms, 'make OT device operator screens lie'; Nvidia found $1B under the couch to help secure American scientific computing dominance Commitment comes as GPUzilla prepares to fortify Uncle Sam's arsenal with at least seven AI-optimized supers; TSMC taps GlobalFoundries to bolster US silicon interposer production in $2B deal Advanced packaging tech is essential to the domestic production of high-performance semiconductors used in AI datacenters

Infosec

Security Russians are posing as Signal support to launch phishing attacks PLUS: US takes down Iranian propaganda sites; Marketing company asks 'Why Do We Have Your Information?' And more!; Security Microsoft patches failed to fix on-prem SharePoint, which is now under zero-day attack PLUS: China upgrades smartphone surveillance tools; Ring eases anti-snooping stance; and more; Black Hat and DEF CON DEF CON Franklin project enlists hackers to harden critical infrastructure Voting village reports have been so successful, says Jeff Moss, that the whole of DEF CON will now be included; Security EQT buys majority share in Swiss cybersecurity biz Acronis Went at equivalent of $3.5B+ valuation for entire firm, though portion sold not specified; Malware Month Ten years since the first corp ransomware, Mikko Hyppönen sees no end in sight On the plus side, infosec's a good bet for a long, stable career

FOSS

NetBSD may have found a new role: The AI-free FOSS OS Increasing awareness and adoption, followed by increasing donations; OpenAI dots inspire open source imitators amid technical difficulties Persistent agents appeal to AI companies, but demand isn't obvious; Ubuntu 'Stonking Stingray' beta swims out: Deeper Rust 'oxidization' plus LLM speech-to-text GNOME 51 is here, and Ubuntu interim releases are getting interesting again; Revive Raskin's 3 laws of software with humane work that considers a dev's duty of care All this, and angry NeoVim users; Firefox 157: Could bold new look be just what Mozilla needs?* Yet another new version of Firefox, with a significant reskin; KDE turns 30 and someone's brought an AI-native desktop proposal Akademy talk imagines Plasma assembling itself around a personal model of each user