by admin | Sep 10, 2026 | Research & Productivity
If you spend your evenings buried in PDFs for a literature review, you already know the real problem is not finding papers. It is finding the time and the eyesight to actually get through them. Zotero, the free reference manager millions of students and researchers already use, just shipped something aimed straight at that problem.
The tool released Zotero 9 this month, and the headline feature is called Read Aloud. It uses natural, AI-generated voices to read your PDFs, ebooks, and saved web pages back to you. It is a small addition on paper, but if you use Zotero for your research and productivity routine, it changes how you can work through a stack of papers.
What is the Zotero Read Aloud feature?
According to Zotero’s own release notes, Read Aloud reads your documents to you in “high-quality, natural-sounding voices,” and it works across PDFs, EPUBs, and webpage snapshots you have saved into your library. You start it with a headphones button in the reader toolbar, and from there you can skip forward or backward by sentence or paragraph, or jump straight to a section by clicking in the margin.
Your place in the document is saved and synced, so you can start listening on your laptop during breakfast and pick up on the same sentence later on another device. There is also an “Annotate Sentence” shortcut that highlights or underlines whatever line you just heard, which is genuinely useful if you are the type who highlights everything on a first pass anyway.
Free voices vs premium voices
Zotero offers two tiers of what it calls Zotero Voices. Standard voices run on Zotero’s own servers and are available to every account, with a monthly free allowance and unlimited use for paid Zotero Storage subscribers. Premium voices are processed by outside text-to-speech providers, sound noticeably more natural, and support more languages, with a smaller free allowance included for everyone to try them.
Read Aloud needs an internet connection and a free Zotero account to use these AI voices. If you would rather stay offline, you can still use your computer’s built-in text-to-speech voices, though Zotero is upfront that the quality is “significantly degraded” compared with the online voices. For now, the feature lives in the desktop app only, with mobile support planned.
Tip: Try Read Aloud on a paper you have already read once. Listening to a second pass while you glance at figures and tables is a fast way to catch details you skimmed the first time, without opening a new tab or tool.
Other Zotero 9 changes worth knowing
Read Aloud is the headline, but a few smaller changes matter for anyone doing serious research work:
- A new “Recently Read” collection shows the items you opened most recently, so you stop hunting for that one PDF you were reading yesterday.
- You can now insert PDF annotations directly into a Word or Google Docs document, with active citations attached, instead of copying them into a Zotero note first.
- Login now happens through your browser rather than by typing your password into the app, which also opens the door to two-factor authentication.
None of this replaces the AI research tools built on top of Zotero, either. Zotero has no built-in chatbot or paper-summarizing AI, and the project says as much in its own plugin documentation: community plugins remain the way people add that kind of feature. If you want an AI assistant that can chat about your library or help you summarize research papers with AI, that still comes from a third-party plugin you install yourself, not from Zotero directly, so only install plugins from developers you trust.
What this means for your research routine
From my own experience working on websites and digital tools, the features that actually stick are the boring, practical ones, not the flashiest ones. Read Aloud fits that pattern. It will not write your literature review for you, and it will not fix a bad paper. What it does is turn dead time, a commute, a walk, folding laundry, into reading time you would otherwise lose.
If you are working through a long reading list, that adds up. Pair it with the habit of double-checking anything an AI tool tells you: text-to-speech is low-risk, but AI summarizers and chat assistants can still invent a citation that does not exist, so always confirm sources against the original paper before you rely on them in your own writing.
Common Questions
Is Zotero Read Aloud free to use?
Yes. Every Zotero account gets some free Standard voice minutes each month, and a smaller free allowance of Premium voices to try. You do not need a paid subscription to use the feature, though heavy use of Premium voices may need a Zotero Storage subscription over time.
Does Read Aloud work on mobile?
Not yet. Zotero says it is currently available only in the desktop app, with iOS and Android support planned for a future update.
Is Zotero’s AI voice the same as an AI chatbot for research?
No. Read Aloud only converts text to speech. It does not summarize, answer questions, or generate citations. Zotero’s own documentation says AI chat and summary features still come from third-party community plugins, not from Zotero itself.
Do I need to update Zotero to get this feature?
Yes. Read Aloud shipped with Zotero 9, so you will need to update your app through Help > Check for Updates, or download the newest version, to see it.
Final takeaway
Zotero Read Aloud will not change how you cite a source or how you argue a thesis. What it does is quietly give you back reading time you were losing, using AI voices good enough to actually listen to. If you already use Zotero, updating and trying it costs nothing but a few minutes. If you are still managing your references in a messy folder of PDFs, this is as good a reason as any to finally set Zotero up properly.
by admin | Sep 6, 2026 | AI Guides
You have probably seen the headlines. One expert says human-level AI is only a few years away. Another says it is decades off, if it ever arrives at all. In the middle of all that noise sits one three-letter word: AGI. If you have ever wondered what it really means, and whether you should be excited or worried, this guide is for you.
AGI stands for artificial general intelligence, and it is one of the most talked-about ideas in technology right now. The tricky part is that it does not exist yet, and the people building AI cannot even agree on what it would look like. Let us break it down in plain English, without the hype.
What is AGI, exactly?
So what is AGI? Artificial general intelligence is a hypothetical future stage of AI where a single system could match or beat humans across almost any mental task. Not one narrow job, but the full range: reading, reasoning, planning, learning something brand new, and moving between very different problems the way a person can.
IBM describes AGI as a hypothetical stage where an AI system can match or exceed human cognitive abilities across any task. The key word there is hypothetical. As of today, no AI system comes close, and there is genuine debate about whether current methods can ever get us there. If you want a refresher on the basics first, our simple explanation of what AI is covers the ground.
Narrow AI vs general AI: the real difference
Almost every AI you use today is what researchers call narrow AI. It is very good at one thing, or a small set of related things, and lost outside that lane.
Think about the examples we already have. A chess engine can beat any human world champion but cannot write you an email. Systems that predict protein shapes do it better than any scientist but cannot drive a car. Even the most advanced chatbot is, underneath, a specialist at predicting text.
General intelligence is different. A general system would carry what it learns in one area over to a completely new one, without being retrained from scratch. That flexible, transfer-it-anywhere ability is exactly what today’s AI lacks. If you want the bigger picture on how these systems relate, our guide on AI vs machine learning vs deep learning walks through the layers.
From my own experience building websites and using AI tools most days, this narrow-versus-general gap is easy to feel. The tools are genuinely useful, but each one is a specialist. Ask it to step outside its training and it quietly falls apart.
Why experts cannot agree on what AGI even means
Here is the part that surprises most people: there is no agreed definition of AGI. The term was popularised back in 2007, and researchers have been arguing over it ever since.
OpenAI, in its official charter, defines AGI as highly autonomous systems that outperform humans at most economically valuable work. Google DeepMind researchers took a different route in a 2023 paper, proposing levels of AGI, a bit like the levels used for self-driving cars, so progress can be measured instead of argued over as a single finish line.
Even Sam Altman, the head of OpenAI, has admitted the word has become a very sloppy term. When the people racing toward a goal cannot describe the goal, you know it is genuinely hard to pin down.
Quick tip: When you see a headline claiming a company has reached AGI, check which definition they are using. Because there is no single agreed test, almost anyone can claim it, and almost anyone can deny it.
Are ChatGPT and other AI tools already AGI?
Short answer: no. The large language models behind tools like ChatGPT are impressive and broad, but they are not general intelligence in the full sense.
They still make things up, they have no real memory of you between chats unless a feature is switched on, and they cannot learn a brand-new skill on their own the way a child can. As Meta’s chief AI scientist Yann LeCun has pointed out, today’s models lack common sense, real-world understanding, and the ability to plan before they act.
The same goes for AI agents, which can take actions and finish multi-step tasks. They are a real step forward, but they still work inside defined limits. Powerful and narrow is not the same as general.
Is AGI the same as superintelligence?
Not quite, and the difference matters. AGI would be roughly human-level across the board. Artificial superintelligence, often shortened to ASI, would go far beyond the smartest humans at nearly everything.
You cannot have superintelligence without first reaching general intelligence, but you can imagine an AGI that is simply as capable as an average person and no more. Most of the dramatic science-fiction scenarios are really about superintelligence, which is even further away and even more hypothetical.
When will AGI arrive?
Nobody knows, and anyone who says they know for certain is guessing. In one large 2023 survey of thousands of AI researchers, the middle estimate for a 50% chance of machines outperforming humans at every task landed around the year 2047. Others are far more hopeful, or far more cautious.
Forecasts like this have a poor track record. The Wright brothers themselves once predicted human flight was 50 years away, only two years before they flew. The honest answer is that AGI could be a decade away, several decades away, or may need a breakthrough we have not discovered yet. The newest models, like GPT-5, are remarkable, but a bigger chatbot is not automatically a general mind.
What AGI actually means for you right now
Here is the practical part. AGI does not exist today, so you do not need to plan your life around a machine that can do everything. What does exist is a growing set of narrow AI tools that are already useful for study, work, and everyday tasks.
The smartest move is not to wait for some future super-AI. It is to get comfortable with the tools we have now: learn to write good prompts, understand what these systems can and cannot do, and stay a little sceptical of anything they tell you. Those habits will serve you no matter what arrives later.
One more thing from working around cybersecurity and online tools: a lot of hype, and a fair number of scams, ride on big words like AGI. If a product promises AGI-powered miracles, slow down and check what it actually does.
Common Questions About AGI
Does AGI exist yet? No. Every AI system available today is narrow AI, strong at specific tasks but unable to match human flexibility across all of them.
Is AGI dangerous? AGI does not exist, so it poses no danger today. Researchers do study future safety, but current real-world concerns are more about how narrow AI is used, such as scams, bias, and misinformation.
What is the difference between AGI and ASI? AGI would be roughly human-level across tasks. ASI, artificial superintelligence, would far exceed humans. AGI would come first.
Will AGI take everyone’s jobs? Since AGI does not exist, this is speculation. For now, the practical shift is learning to work alongside today’s narrow AI tools.
Final takeaway
AGI is a big, exciting idea, but it is still a goal rather than a reality. Today’s AI is powerful and narrow, not general. Instead of worrying about a machine that can do everything, focus on understanding and using the very capable tools already in front of you. That is the real way to prepare for whatever comes next.
by admin | Sep 5, 2026 | Opportunities
You finished a free AI course, you can write a decent prompt, and then a job listing asks for “proven experience with AI.” It’s a frustrating wall. How are you supposed to prove you can do the work when nobody has hired you to do it yet?
The honest answer is that you build the proof yourself. An AI portfolio is simply a small, public collection of things you have made that anyone can open with a link. It matters more than most people expect, because a certificate says you sat through a course, while a project shows you can actually use what you learned. And you can build a solid AI portfolio for free, with no job, no degree, and no expensive tools.
Why an AI portfolio beats a certificate
Certificates are useful, but they mostly prove attendance. A project proves ability. When someone can click a link and see a thing you built, the conversation changes from “trust me” to “look what I did.” Employers and clients trust what they can try for themselves.
From my own experience building websites and small online tools, the thing that got people to take me seriously was never a line on a CV. It was something they could open in a browser and use. A portfolio does the same job for AI skills.
What goes into an AI portfolio?
You don’t need much to start. A strong beginner AI portfolio usually has:
- Two or three small projects you actually finished.
- A short write-up for each one: what problem it solved, what you did, and what you learned.
- Links people can click, such as a notebook, a live demo, or a code repository.
- Optional badges from free courses, kept in their place.
Important tip: one finished project you can explain clearly beats ten half-built ones. Depth is what people remember, not the length of the list.
Start with small, real projects
Pick something you actually care about, then keep it tiny. A good first project can be as simple as taking a public dataset (movie ratings, weather, sports stats) and finding one interesting pattern in it. Or build a custom assistant for a hobby you know well. If you want a gentle starting point, our guide on how to make your own AI assistant for free walks through a project you can finish in an afternoon.
And no, you do not need to be a programmer. Plenty of real projects use no-code tools, prompt libraries, or a clear written analysis. If code feels like a barrier, start with our guide on how to learn AI without coding, then come back and build. If you still need the basics, learning AI for free is a good first step.
Use free platforms to show your work
Once you have made something, it needs a home online. These are all free and widely respected:
- Kaggle: free notebooks that run in your browser, thousands of open datasets, and a public profile. See Kaggle.
- GitHub: free public repositories for any code or files, where a clear README acts as your project’s front page. See GitHub.
- Hugging Face Spaces: when you are ready to go further, you can host a live demo people can actually try, on a free public link. See Hugging Face Spaces.
One habit worth keeping from the security side of my work: only use open datasets or your own data, and never put private, personal, or client information into a public project. Check a dataset’s license before you publish anything based on it.
Enter free competitions for practice and proof
Competitions give you two things at once: a finished project and a story to tell about it. Kaggle’s “Getting Started” competitions, like the well-known Titanic challenge, are built for beginners and never close, so there’s no pressure to rush. Free hackathons and student contests work the same way. We rounded up the best beginner-friendly options in our guide to AI competitions for beginners.
Add free course badges the right way
Badges are a nice supporting detail. On Microsoft Learn, finishing free learning paths earns badges and trophies. Google AI Essentials gives a shareable digital badge through Credly, and Kaggle Learn hands out free certificates too.
Just remember the order of things: a badge supports a project, it does not replace one. Put one or two on your profile, then let your actual work do the talking.
Write up each project so people understand it
The write-up is half the value, and it’s the part most people skip. A recruiter may never run your code, but they will read three clear sentences. For each project, answer three questions: what problem were you solving, what did you build, and what is one thing you learned or would improve next time?
Keep it honest. Don’t claim more than you did. A small, truthful project explained well beats a big one you can’t really talk about.
A simple 4-week plan to start
- Week 1: learn a little and pick one small project idea.
- Week 2: build the smallest version that actually works.
- Week 3: put it on Kaggle or GitHub with a clear write-up.
- Week 4: enter a beginner competition or start project two, then add the link to your CV.
When you do start applying, your portfolio becomes your strongest line. Our guide on how to use AI to apply for jobs shows how to point employers straight to it.
Common Questions
Do I need to know how to code to build an AI portfolio?
No. No-code tools, custom assistants, prompt libraries, and clear written analyses all count. Coding helps for some projects, but a well-explained no-code project is still real proof of skill.
How many projects do I need?
Two or three finished, well-explained projects are plenty to start. A clear write-up matters far more than the number of projects.
Where should I host everything?
Pick one home base people can visit. A GitHub or Kaggle profile works well, and you can add a live demo on Hugging Face Spaces later. Then link to it from your CV and LinkedIn.
Is it safe to put my projects online?
Yes, as long as you only use open or your own data. Never upload private, personal, or client information, and read a dataset’s license before sharing results based on it.
Final takeaway
You don’t need permission or a paycheck to start proving your AI skills. Pick one small idea this week, build the simplest version that works, write three honest sentences about it, and put it somewhere with a link. Do that a few times and you’ll have something most applicants never bother to make: real proof you can use AI, not just talk about it.
by admin | Sep 4, 2026 | AI Tools
You open ChatGPT, explain who you are, what you do, and the tone you want. The answer comes back good. Three days later you open a fresh chat and type the same background all over again.
That loop is the quiet reason a lot of people decide AI tools are more trouble than they are worth. The fix is not a cleverer prompt, it is a ten minute setup you do once. This guide shows you how to make your own AI assistant, for free, using a tool you probably already have open, no coding required.
What it really means to make your own AI assistant
When people hear “build your own AI” they picture training a model on a mountain of data. That is not this: you are not building anything from scratch, and you are not touching Python.
You are creating a saved shortcut with a name, a set of standing instructions, and often a few files attached. Every time you open it, the assistant starts with that background already loaded, like briefing a new colleague properly once instead of reintroducing yourself every morning.
This is not the same thing as an AI agent, which goes off and completes tasks on its own. A custom assistant does not act for you, it just stops forgetting who you are. Our guide to AI agents explained simply covers that difference.
Why a saved assistant beats a very long prompt
You could paste your background at the top of every chat. Plenty of people do, but it has three problems.
- You have to remember to do it, and you will not on the day you are rushed.
- Long background eats into the space the model has to work with, a limit we explained in what is a context window. Spending that room on the same introduction every time is waste.
- Small differences creep in. You describe yourself slightly differently each time and the answers drift with you.
A saved assistant fixes all three at once. The instructions sit outside the conversation, they are identical every time, and any files you attach stay attached.
Option 1: ChatGPT Projects
OpenAI calls these Projects, described in its help pages on Projects in ChatGPT as workspaces that keep chats, files and custom instructions in one place. They are available to all free and paid subscription types globally, so you do not need to pay to try one, just be signed in.
- Click New project in the sidebar and give it a name.
- Open the three dot menu, choose Project settings, and write your instructions there. OpenAI notes these apply only inside that project and override your global custom instructions while you are in it.
- Upload your reference files. OpenAI lists the limit as five files per project on a free account, twenty five on Go and Plus, and forty on Pro, Business, Edu and Enterprise.
Projects also keep their own memory of the chats inside them, and a setting can switch a project to project only memory if you would rather it ignored everything else in your account.
Option 2: Gemini Gems
Google’s version is called a Gem. Its getting started guide for Gems describes them as customised versions of Gemini for repetitive tasks or deep expertise in new areas. The only requirements listed are a personal Google Account and being 13 or over, or the applicable age in your country. No paid plan is required.
- In the Gemini web app, click Explore Gems, then New Gem.
- Name it and write the instructions.
- Under Knowledge, add files from your device or from Google Drive.
- Click Save. Google specifically warns that trying your Gem in the preview panel does not save it, which catches a lot of people out.
Two extras help if you are stuck: a button that asks Gemini to rewrite and expand your instructions, and premade Gems (Brainstormer, Coding partner, Writing editor) you can open and copy the pattern from.
Option 3: Claude Projects
Anthropic’s help page on projects, updated in July 2026, says projects are available to all users including free accounts, and that free users can create a maximum of five of them.
A Claude project is a self contained workspace with its own chat history and knowledge base: you upload documents, set project instructions, and every chat inside starts from that base. The heavier retrieval feature for very large amounts of uploaded material is limited to paid plans, and sharing with colleagues is a Team and Enterprise feature.
If you are not sure which of the three to start with, the honest answer is whichever one you already have open. Our ChatGPT vs Gemini vs Claude comparison goes into how they differ in day to day use.
How to write instructions that actually work
This is where most custom assistants fall down: people write “be helpful and professional”, then wonder why nothing changed.
Google’s own guidance on writing Gem instructions splits the job into four parts, and the same four work just as well inside ChatGPT and Claude:
- Persona. The role it should play and how it should speak to you.
- Task. What it should actually produce.
- Context. Your background, constraints and audience, and anything it would otherwise guess wrong.
- Format. Length, structure, bullet points or prose, and what to leave out.
You do not need all four. Two or three specific ones beat a paragraph of vague adjectives every time, the same habits that make a single prompt work. Our guide on writing better AI prompts applies directly here.
A short example you can copy
Say you write lab reports as a final year student. Rather than explaining that every week, the instructions might read:
You help me write and revise undergraduate biology lab reports. Use plain, precise academic English and British spelling. Never invent citations or results. When I paste a draft, point out unclear reasoning and missing method detail first, before you comment on grammar. Do not rewrite my text unless I ask you to. Give feedback as a short numbered list.
Attach your module handbook and marking rubric as files, and every chat in that project already knows the rules you are marked against. You never type any of it again.
What to keep out of your assistant
From my own experience working with websites, online tools and cybersecurity, this is the part people skip. A file you upload is not a one off. It sits there for every future chat until you delete it, and anyone you later share the project with can usually read it too.
Keep out anything you would not want sitting in a cloud account for months: passwords and API keys, ID or passport scans, client contracts you do not have the right to share, medical records, and unpublished confidential work. OpenAI also notes that on Free, Plus and Pro accounts, project information may train its models if “Improve the model for everyone” is switched on, so check that toggle before uploading anything sensitive.
Important tip: before you attach a document, ask yourself one question. Would I be comfortable if this file were still sitting in this account in a year, readable by anyone I ever share the project with? If the answer is no, paste in only the paragraph you actually need instead of the whole file.
Our guide on using AI safely and protecting your privacy goes further on the settings worth changing before you upload anything.
Common Questions
Do I need to pay to make my own AI assistant?
No. OpenAI says Projects are available to all free and paid subscription types, Anthropic says projects are available to all users including free accounts with a limit of five, and Google’s Gems guide lists a personal Google Account and a minimum age rather than a paid plan. Paid tiers mainly raise the limits, such as how many files you can attach.
Is this the same as training my own AI model?
No. You are not changing the underlying model at all. You are saving instructions and files that get applied at the start of each chat. The model doing the work is the same one everyone else is using.
How many assistants should I make?
Fewer than you think. Two or three matching things you genuinely do every week will get used, twelve will not. Start with one and only add another when you catch yourself typing the same background again.
Will my assistant remember what I told it in an earlier chat?
That depends on the tool and its settings. ChatGPT projects keep their own memory of chats inside them, and you can set a project to ignore everything outside it. Treat the standing instructions and uploaded files as the reliable part, and anything mentioned in passing as the part that may not carry over.
Can I share my assistant with someone else?
Sometimes. OpenAI allows project sharing across its plans, with the number of collaborators depending on the plan. Anthropic limits project sharing to Team and Enterprise accounts. Before you share anything, remember that everyone you invite can usually see the files you uploaded.
Final takeaway
Making your own AI assistant is not a technical project. It is ten minutes of writing down what you would tell a new colleague on their first day, saved somewhere the tool reads it every time. Pick the chat tool you already use, create one assistant for the task you repeat most, and edit the instructions properly after a week of real use. That one setup will save you more time than any prompt trick you will ever read.
by admin | Sep 3, 2026 | Research & Productivity
It is 11pm, your draft is finished, and nobody you trust is awake to read it. So you paste the whole thing into a chatbot and type the four words most of us have typed at some point: “can you improve this?” A few seconds later you get back something smoother, tidier, and slightly stranger. It reads well. It just does not sound like you anymore.
There is a better way to use these tools, and it comes down to one small change. Ask for feedback instead of a rewrite. Good AI feedback on your writing works a bit like a patient reader who never gets tired: it can tell you where your argument wobbles, where a paragraph loses the thread, and what a reader might still be confused about after finishing. Then you decide what to change, in your own words.
This guide covers what to ask for, the prompts worth copying, what to ignore, and the rules you need to check first.
Why AI feedback on your writing beats “make this better”
When you ask an AI tool to improve a text, it will do exactly that, and it will usually do it by replacing your sentences with its own. You end up with a document you did not write and cannot fully defend if somebody asks you why you made a particular choice.
The Writing Center at the University of North Carolina at Chapel Hill puts the principle neatly in its handout on generative AI in academic writing: when instructors allow these tools, they assume the tools will help you think and write, not think or write for you. Feedback keeps you in the driving seat. A rewrite quietly takes the wheel.
There is a practical benefit too. Feedback teaches you something you can reuse next time. A rewrite teaches you nothing except how to press the button again.
Check the rules before you paste anything in
This step takes two minutes and saves a lot of trouble. Course policies differ wildly, sometimes between two modules in the same department, so read the assignment brief before you open a chat window.
Elon University’s Center for Writing Excellence asks students to check the policy for each assignment and disclose how they used AI when they bring a draft in for help, and it advises keeping a record of your prompts so you can explain your process later. That last part is good advice for anyone, student or not. If you cannot describe how a tool helped you, you probably leaned on it too hard.
Privacy matters just as much. From my own work building websites and handling client material, the rule I stick to is simple: if a document would embarrass me or somebody else in a leak, it does not go into a public chatbot. Unpublished research, confidential reports, anything with personal data in it, and anything covered by an employer agreement all belong in that category. Our guide to using AI safely and protecting your privacy walks through the settings that reduce the risk.
Give the tool the brief, not just the text
Most disappointing feedback comes from a missing brief. The tool cannot judge whether your essay answers the question if it has never seen the question.
Before you ask for comments, paste in four things: the actual task or assignment wording, the marking rubric or the criteria your reader cares about, who the audience is, and the word limit. Then add your draft. The difference in quality is dramatic, because the tool now has a standard to measure your work against instead of a vague idea of “good writing”.
The study skills team at the University of Roehampton also suggests breaking your prompts down and asking for one thing at a time, listing spelling and grammar, structure, argumentation and reasoning, and flow as separate jobs. Ask for all four at once and you get a shallow paragraph on each. Ask for one and you get something you can act on. If prompt wording is where you usually get stuck, our guide on writing better AI prompts covers the basics.
Five feedback prompts worth copying
These are written so the tool comments rather than rewrites. Notice that none of them say “fix” or “improve”.
- Structure: “Here is my assignment brief and my draft. Do not rewrite anything. List the main point of each paragraph in one line, in order, so I can see whether my structure makes sense.”
- Argument: “Act as a sceptical reader. What is my main claim, and where is the evidence weakest or missing? Quote the exact sentences that worry you.”
- Clarity: “Mark any sentence a tired reader would need to read twice, and say in a few words why. Do not suggest replacements yet.”
- Gaps: “What questions would a reader still have after finishing this? What did I assume they already know?”
- Voice check: “Describe the tone of this draft in three phrases. Does it match a piece written for [your audience]?”
The quoting instruction in the second prompt matters more than it looks. It forces the tool to point at real sentences in your text instead of producing generic advice that would fit any essay ever written.
Important tip: ask for the reasoning behind every suggestion, then decide for yourself. If the tool cannot explain why a change helps, treat the suggestion as an opinion and move on.
What to do with feedback you disagree with
You are allowed to say no. Some AI suggestions are wrong for academic or professional writing in specific, predictable ways, and the Roehampton guide flags two of them. The first is a push towards first person and direct personal pronouns, which can make writing clearer but can also weaken the tone your marker expects. The second is a habit of strengthening hedged language. Phrases like “the evidence suggests” exist for a reason, and turning them into confident claims can push you past what your data actually supports.
There is a useful exercise in the UNC handout for spotting this. Run a paragraph through the tool, then put the original and the edited version side by side and look at what changed. Do you agree with the changes? Did the edit make the meaning clearer, or did it just sand off the parts that sounded like you? Doing this a few times teaches you a lot about your own habits, which is the real prize here.
Where AI feedback goes wrong
Three failure modes come up again and again, and all three are worth watching for.
- Confident wrong facts. If you ask a tool to check a claim or suggest supporting sources, it may invent them. UNC’s handout warns that citations can look perfectly formatted and still be fictional. We covered how to catch this in our post on checking AI citations.
- Flattening. Rounds of “smoother” edits tend to converge on the same neutral house style. Two or three passes and your writing sounds like everyone else’s.
- Doing the thinking for you. The most comfortable failure mode. Accepting feedback without judging it is still outsourcing the work, which is exactly what our piece on AI and critical thinking digs into.
A simple routine that takes about twenty minutes
Finish your draft first, without any AI in the room. Then paste in the brief and the rubric, and run the structure prompt. Fix the structure yourself. Run the argument prompt next, and note which weaknesses you accept. Fix those yourself too. Only at the end, once the thinking is settled, ask about clarity and grammar at sentence level.
Elon’s guidance makes the same point about timing: used carefully, AI helps most at the beginning of the writing process, during brainstorming, and at the end, during proofreading. The middle, where you work out what you actually think, is yours. Keeping that middle for yourself is also the honest answer to the academic integrity question, which we look at in more detail in our guide to using AI tools as a student without cheating.
Common Questions
Is asking AI for feedback on my writing cheating?
It depends entirely on your institution and the specific assignment, which is why you check the brief first. Many universities treat feedback and proofreading differently from generating text, but policies vary by module and by tutor. If the brief is unclear, ask before you submit, not after.
Which AI tool gives the best writing feedback?
The prompt matters far more than the tool. The main free chat assistants all handle this task reasonably well. What changes the outcome is whether you supplied the brief and the rubric, and whether you asked for comments rather than a rewrite.
Should I paste my whole document in at once?
For structure feedback, yes, because the tool needs to see the shape of the whole piece. For sentence level work, go section by section. Feedback on a short passage is almost always more specific and more useful than feedback on twenty pages.
How do I stop AI feedback from changing my voice?
Never paste the suggested version back into your document. Read the comment, close the tab, and make the change in your own words. If a suggestion is good, you will remember it well enough to write it yourself.
Final takeaway
Used as an editor, AI takes over your writing. Used as a reader, it makes you better at it. Give it the brief, ask for comments instead of corrections, check the rules for your course or your workplace, and make every change yourself. Try one prompt from the list above on your next draft and see how much of the feedback you actually agree with. That disagreement is where your own judgement lives, and it is worth protecting.
