by admin | Jul 9, 2026 | Research & Productivity
Ask anyone who has written a thesis which part they underestimated, and you will usually get the same answer: the literature review. You start with one search, and two weeks later you have sixty open tabs, a folder full of unread PDFs, and no clear picture of the field.
AI tools can remove a lot of that pain if you point them at the right jobs. This guide walks through a five step literature review with AI, using free tools, and it stays honest about the parts you still need to do yourself.
Can you really do a literature review with AI?
Partly, yes. AI is genuinely good at three jobs here: finding papers that match your question even when you do not know the perfect keywords, summarizing individual papers quickly, and showing how papers connect to each other.
What it cannot do is judge research quality the way you can, decide why a gap in the field matters, or build your argument. AI models also make mistakes with total confidence. They sometimes invent references or misread a paper\u2019s findings, a problem we explained in AI hallucinations explained. So the workflow below uses AI for speed and keeps the judgement with you.
Someone close to me spends her days in PhD research on machine learning and medical imaging, so I have watched how fast a reading pile can grow. The researchers who cope are not the ones reading faster. They are the ones with a better system.
Step 1: Turn your topic into a real question
\u201cAI in healthcare\u201d is a topic. \u201cHow accurate are deep learning models at detecting brain tumours from MRI scans?\u201d is a question. Every step that follows works better when you start from a question, because modern research tools use semantic search. They match meaning, not just keywords.
Write your question down before you open any tool. If you cannot phrase it yet, that is useful information too. Spend an hour with a general overview or a textbook chapter first, then come back.
Step 2: Find papers with AI search tools
Three tools cover most of the discovery work:
- Elicit searches more than 138 million papers. You type your question and it returns a table of relevant papers with short summaries. The Basic plan is free, and it can import your library from Zotero.
- Semantic Scholar is a free academic search engine from the non-profit Allen Institute for AI. It indexes over 200 million papers and adds short AI generated summaries, called TLDRs, so you can screen results quickly.
- Research Rabbit maps papers visually. You start with one paper you already trust, and it shows similar, earlier, and later works, so you follow the citation trail instead of searching blind.
University libraries have started recommending these tools too. The University of Michigan Library keeps a practical guide on AI in literature reviews if you want a librarian\u2019s take on the same tools.
Tip: run the same question through two different tools. Each one searches differently, and the papers that appear in both lists are usually the ones to read first.
Step 3: Screen and organize what you find
You will collect far more papers than you need, so do not try to read them all. Screen each one by its abstract or TLDR and sort it into three piles: keep, maybe, and drop. Be ruthless with the drop pile.
For the keepers, use Zotero, a free and open source reference manager. It stores your citations, formats them in thousands of styles, and connects with Elicit and Research Rabbit. We covered where it fits in our guide to AI tools for thesis writing.
Step 4: Summarize and compare the papers
For every paper you kept, you want four things: the question it asked, the method it used, what it found, and its limitations. AI can speed this up a lot. Our guide on how to summarize research papers with AI shows practical prompts, and our NotebookLM and Elicit walkthrough covers tools that answer questions only from the sources you upload.
One warning from experience: AI extraction makes mistakes. It can misread a sample size or blur two findings together. Check every number and claim against the actual paper before it goes anywhere near your draft.
Step 5: Write the review yourself
Here is the part no tool can do. A literature review is not a list of summaries. It is an argument about the state of a field: what researchers agree on, where they clash, and which gap your work will fill. That structure has to come from your reading, so group your papers by theme or debate, not by author.
Two rules protect you here. First, verify that every reference exists and says what you claim, because AI generated citations are sometimes fake. Second, check your university\u2019s AI policy and disclose what you used. Most universities now allow AI for searching and summarizing but treat AI written text as misconduct.
From my own work with websites and online tools, the pattern is always the same. Tools that remove boring steps earn their place. Tools that promise to think for you cause trouble later.
Common Questions
Can AI write my literature review for me?
It can produce text that looks like one, but that is the trap. The references may not exist, the synthesis is shallow, and most universities treat submitting it as academic misconduct. Use AI to find, organize, and summarize. Write the argument yourself.
Are these AI research tools free?
Yes, for everything in this workflow. Semantic Scholar is completely free, Elicit has a free Basic plan, Research Rabbit lets you sign up free, and Zotero is free and open source.
How many papers should a literature review include?
It depends on your field and level. A bachelor\u2019s thesis might cover 20 to 40 papers, while a PhD literature review can pass 150. Your supervisor\u2019s guidance beats any general number, so ask early.
Final Takeaway
A literature review with AI is not about outsourcing the reading. It is about shrinking the boring parts: hunting for papers, formatting citations, and writing first pass summaries. Pick one question, run it through Elicit or Semantic Scholar this week, and save what you find into Zotero. The pile gets smaller, the map gets clearer, and the thinking stays yours.
by admin | Jul 8, 2026 | Future Jobs
You click the interview link, fix your hair in the little preview window, and wait for someone to join. Nobody does. Instead, a recorded voice reads out the first question and a timer starts counting down. Your interviewer today is an AI.
If that sounds unusual, it isn’t anymore. According to the Greenhouse 2026 Candidate AI Interview Report, published in May 2026, 63% of job seekers have already been interviewed by an AI. That number jumped 13 points in just six months. So if you’re applying for jobs this year, the real question is not whether you’ll face an AI job interview. It’s when, and how ready you’ll be.
What is an AI job interview?
An AI job interview is any interview where software, not a person, asks the questions or scores your answers. It usually takes one of three forms. The most common is the one-way video interview: you record answers to preset questions and an algorithm (sometimes with a human reviewer) rates them later. Some companies use chat-based screenings, where you type answers to a bot. And a growing number now use live AI voice interviewers that ask follow-up questions in real time.
Employers like these tools because they can screen thousands of applicants quickly. For you, it means the first “person” standing between you and the job is often a piece of software.
What the AI is actually measuring
This is the part most candidates never get told. The Greenhouse survey found that 70% of job seekers were never clearly informed that AI would evaluate them, and 39% said they want employers to explain what the AI measures.
In practice, most systems look at the content of your answers: the skills you mention, how closely your language matches the job description, and how clearly you structure your response. Duke University’s career guidance for video interviews makes this point directly: study the job description, pick out the key skills and qualifications, and work them naturally into your answers, because that’s largely what the software is listening for.
Why so many candidates walk away
AI interviews have a trust problem right now. In the same survey, 38% of candidates said they had abandoned a hiring process because it included an AI interview. The biggest complaints were pre-recorded video interviews scored with no human present, companies not disclosing how AI would be used, and AI monitoring during the process.
The aftermath can sting too. Of the candidates who completed an AI interview, 51% simply never heard back. That’s worth knowing before you start: silence after an AI screening is common and says very little about you. Don’t take it personally, and don’t stop applying elsewhere while you wait.
How to prepare for an AI job interview
The good news is that AI interviews reward preparation more than charm. Here’s what actually helps:
- Mirror the job description. Reread it before the interview and note the exact skills it asks for. Use those words in your answers where they’re true for you.
- Structure every answer. The STAR method (Situation, Task, Action, Result) keeps your response clear for both algorithms and humans.
- Answer first, then explain. Get to the point in your first sentence or two, then back it up with a real example.
- Practice out loud. Record yourself answering two or three common questions on your phone. You’ll hear the rambling immediately.
- Don’t read from a script. Your eyes give it away on camera. A few keyword notes near the screen are fine; a full script is not.
From my own work building websites and running digital projects, I’d add one thing people always underestimate: test your tech. Check your camera, microphone, lighting, and internet connection the day before, the same way you’d test a website before launch. A frozen video or muffled audio can sink a good answer, and it’s completely avoidable.
Tip: treat the practice questions seriously. Most AI interview platforms offer a test question before the real one starts. Use it to check your sound and framing, not just to relax.
What you’re allowed to ask the employer
Asking about AI in the hiring process is reasonable, and increasingly normal. In the full Greenhouse report, 57% of candidates said disclosure should be a legal requirement, and 46% want the option to request a human interview instead. Some employers already offer one.
Before the interview, it’s fair to ask three things: Will AI be used to evaluate me? What does it measure? And will a human review the result before a decision is made? A company that answers openly is telling you something good about how it treats people.
There’s a privacy side too. These platforms record your video and voice, so it’s smart to check how long recordings are kept. I write about this mindset in my guide on how to use AI safely and protect your privacy, and it applies here as well.
Should you use AI to prepare?
Yes, for practice. Ask ChatGPT or a similar tool to act as an interviewer for your specific role, then answer its questions out loud. It’s one of the most useful tricks in our guide to using AI in your job search. What you shouldn’t do is have AI feed you live answers during the interview. Detection aside, you’d be rehearsing for a job with someone else’s voice.
And keep the bigger picture in mind. AI is changing interviews because it’s changing work itself. If you want to understand where that’s heading, our posts on whether AI will take your job and why AI skills now pay more cover what the data really says.
Common Questions
Do employers have to tell me an AI is interviewing me?
In most places, not yet, and in practice most don’t: 70% of candidates in the Greenhouse survey were never clearly told. Rules are tightening in some regions, but for now the safest move is simply to ask before the interview.
Can I refuse an AI interview?
You can always ask for a human alternative, and 46% of candidates want that option as standard. Some employers will say yes. If they won’t, you can still decide whether the role is worth it, which is exactly what 38% of candidates have done.
How do I know if the AI rejected me?
Often you won’t. Only 13% of candidates in the survey were formally rejected after an AI interview, while 51% never heard back at all. Follow up once after a week or so, then keep applying elsewhere.
Final takeaway
AI job interviews are now a normal part of getting hired, even though most of them still aren’t handled well. You can’t control whether a company uses one, but you can control how prepared you are: know what the software measures, structure your answers, test your setup, and ask honest questions about how AI is being used. Do that, and the AI screening stops being scary. It becomes just one more door on the way to the job.
by admin | Jul 7, 2026 | AI Guides
Have you ever pasted a long document into an AI chatbot and watched it stop halfway? Or seen an AI plan advertise a “128K token” limit and wondered what that actually means? Tokens sit behind almost everything AI chatbots do, and once you understand them, a lot of confusing AI behaviour suddenly makes sense.
In this guide, you’ll learn what a token in AI is, how your words get chopped into tokens, and why token limits explain the message caps, forgetful chats, and prices you run into every day. Plain English, no math needed. If you’re completely new to this topic, our simple explanation of what AI is is a good place to start.
What is a token in AI?
A token is a small piece of text that an AI model reads and writes. It’s the basic unit every AI language model works with. A token is not the same as a word. It can be a whole short word, part of a longer word, a punctuation mark, or even a space attached to a word.
OpenAI’s help guide gives some handy rules of thumb for English text:
- 1 token is roughly 4 characters
- 1 token is about three quarters of a word
- 100 tokens come to roughly 75 words
So a 1,000 word blog post like this one is somewhere around 1,300 to 1,400 tokens. Numbers, code, and unusual words push the count higher.
How AI turns your words into tokens
Before an AI model ever sees your message, a piece of software called a tokenizer splits it into tokens. Each token then gets an ID number, because the model works entirely with numbers, not letters. Common short words usually stay whole. Longer or rarer words get broken into smaller chunks.
GPT models use a subword method called Byte-Pair Encoding, as Microsoft’s guide to tokens explains. To give you a feel for the scale, OpenAI notes that the famous quote about missing all the shots you don’t take is 11 tokens long.
Language matters too. OpenAI points out that the Spanish phrase “Cómo estás” takes 5 tokens for just 10 characters. Many non-English languages break into more tokens per word, which makes the same request slower and more expensive than it would be in English.
Why do AI models count tokens instead of words?
Because tokens are literally how these models think. A large language model generates text by predicting the next token, one token at a time, then feeding its own output back in and predicting the next one again. Words are for us. Tokens are for the model.
Subword tokens also give models flexibility. When a model meets a brand new word, a typo, or an unusual name, it can still handle it by piecing together smaller chunks it already knows.
What is a context window?
Every AI model has a limit on how many tokens it can handle at once. That limit is called the context window, and it covers your input and the model’s output together. When a conversation grows past it, the oldest parts effectively fall out of view.
This is why a long chat starts to “forget” things you said earlier. The AI isn’t being lazy. Those early messages simply no longer fit inside the window it can see. The practical fix is to start a fresh chat and paste in only the details that still matter, or keep your prompts focused from the start. Our guide on writing better AI prompts helps a lot here.
Why tokens decide what AI costs
Most paid AI services charge by the token, and input tokens are often priced differently from output tokens. Free plans work the same way underneath, they just cap how much you can use. That’s why generative AI tools talk about tokens so much: they measure the actual work the model does.
From my own experience connecting AI tools to websites and small projects, this was the first surprise: the bill counts tokens, not questions. A one-line question is nearly free. The same question with a 20-page document pasted under it costs many times more, because every word of that document becomes tokens the model has to read.
How to see tokens for yourself
The easiest way to make all this real is to try OpenAI’s free Tokenizer tool. Paste in any sentence and it shows you exactly how the text splits into coloured chunks, with a live token count. Two minutes with it teaches you more than any definition.
Quick tip: if an AI chat starts forgetting details or giving weaker answers, your conversation has probably outgrown its context window. Start a new chat and paste in only the information that still matters.
Common Questions
How many words is 1,000 tokens?
Roughly 750 words in English, using OpenAI’s rule of thumb. The exact number depends on your language and word choices, since longer and rarer words split into more tokens.
Do spaces and punctuation count as tokens?
Yes. Spaces usually attach to the word that follows them, and punctuation marks often become their own tokens. Everything you type contributes to the token count.
Why do AI plans talk about tokens instead of messages?
Because messages vary hugely in size. A short question and a pasted 50-page report are both “one message,” but the report takes far more computing work. Counting tokens is the fair way to measure that work.
Final takeaway
Tokens are the small chunks of text every AI model actually reads and writes. They explain the limits on your chats, the reason long conversations drift, and the way AI pricing works. You never have to count them by hand. But once you know they exist, AI tools stop feeling mysterious and start feeling like something you can plan around.
by admin | Jul 6, 2026 | AI Tools
Writing a thesis can feel like doing three jobs at once. You are the researcher hunting for papers, the writer drafting chapters, and the admin keeping hundreds of references in order. So it makes sense that so many students now search for an “AI thesis writer” and hope one tool will do it all.
Here is the honest answer up front: no AI tool should write your thesis, and the ones that promise to are the ones to avoid. But the right tools, used openly and carefully, can save you real hours every week. This guide walks through the AI tools for thesis writing that actually help, what each one is good at, and the rules to check before you touch any of them.
Before the tools: one rule that protects you
Universities now treat AI use in a thesis as something you agree with your supervisor first, not something you quietly do on the side. The University of Toronto’s graduate school guidance, updated in June 2026, is a good picture of where things stand: get clear approval from your supervisor before using generative AI for research or writing, and describe in the thesis which tools you used, how, and why.
Your university will have its own version of these rules, and they can differ between departments in the same building. Check them first, get the agreement in writing, and keep notes on what you used. That one short conversation protects your degree.
What AI tools for thesis writing can honestly do
Think of AI as a research assistant, not an author. It is genuinely useful for four jobs: finding relevant papers, understanding sources faster, keeping citations organised, and sharpening text you wrote yourself.
What it cannot do is produce the original contribution a thesis is judged on. University guidance points out that AI-generated text may not meet originality requirements, and you are fully responsible for anything it produces, including its mistakes. If a chatbot writes a paragraph and that paragraph is wrong, it becomes your problem in the exam room, not the chatbot’s.
Finding papers: Elicit
Elicit is built for academic search. Instead of guessing keywords, you ask a research question in plain English and it searches a database of over 138 million papers, returning answers with citations that link back to the underlying sources. That makes it a strong starting point for a literature review, because you can see what already exists before you commit to a research gap. There is a free version, which is enough to test it on your own topic.
Treat it as a discovery tool. Skim what it surfaces, then read the papers that matter yourself. We covered a sensible reading workflow in our guide on how to summarize research papers with AI.
Understanding your sources: NotebookLM
Once you have a pile of PDFs, Google’s NotebookLM lets you upload them and ask questions that are answered only from those documents, with citations pointing to the exact passages. That grounding makes it far safer for thesis work than a general chatbot, because it works from your sources instead of its memory.
It shines when you need to interrogate your own literature pile. “Which of these papers used a sample under 100 people?” becomes a ten second question instead of an afternoon. Our beginner’s guide to NotebookLM walks through setting it up step by step.
Keeping citations honest: Zotero
Zotero is free, open source, and has been the quiet workhorse of academic writing for years. It collects papers as you browse, organises them into collections, and formats references in over 9,000 citation styles directly inside Word, LibreOffice, and Google Docs.
Why does a citation manager belong on an AI list? Because chatbots are famous for inventing references that look real but do not exist. Every citation in your thesis should come from a paper you actually opened and saved. If you want to understand why AI makes sources up, our explainer on AI hallucinations is worth five minutes.
Improving your writing without losing your voice
The safest way to use ChatGPT, Claude, or Gemini on thesis text is as a critic, not a ghostwriter. Paste a paragraph you wrote (once your supervisor has agreed to this) and ask it to flag unclear sentences, weak transitions, or claims that need evidence. You keep the writing. It supplies the questions. And keep unpublished data out of chatbots entirely; our guide on using AI safely explains what should never be pasted into a free tool.
Quick tip: ask AI to interrogate your draft instead of rewriting it. A prompt like “List the three weakest arguments in this section and ask me the questions an examiner would” makes your thinking better, and the words stay yours.
I have watched PhD researchers around me lose whole evenings to reference lists a free tool could format in seconds. The pattern repeats everywhere, including in my own work with websites and digital projects: AI helps most with the boring jobs and least with the thinking jobs. A thesis is mostly a thinking job.
What to avoid
Be careful with anything marketed as an “AI thesis writer” or “dissertation generator”. A tool that promises finished chapters is selling you an academic misconduct case with a subscription button. Unauthorized AI use can be treated as an offence under university codes of conduct, and examiners can and do ask you to defend every paragraph.
Also, do not cite a chatbot as if it were a source. If AI use is permitted, you disclose the use itself. Style guides now cover this; the APA Style guidance on citing ChatGPT is a widely used example.
Common Questions
Can AI write my thesis for me?
No. A thesis is assessed on your original contribution, and AI-generated work may not meet that bar. Unauthorized AI writing can count as misconduct, and you must defend the text in your oral exam either way.
Do I have to tell my university I used AI?
In most cases, yes. Current university guidance expects supervisor approval in advance plus a clear description in the thesis of which tools you used and how. Rules vary by institution and department, so always check yours.
What is the best free AI tool for thesis writing?
There is no single best tool because the jobs are different. Elicit is strong for finding papers, NotebookLM for questioning your own sources, and Zotero for citations. All three have free versions, and together they cover most of the thesis workflow.
Final takeaway
AI will not write a good thesis, but it can clear the path so you can. Agree the rules with your supervisor, let Elicit widen your reading, let NotebookLM question your sources, let Zotero guard your references, and keep the writing yours. The thesis with your voice in it is the one worth defending.
by admin | Jul 5, 2026 | AI Guides
If you’ve ever asked ChatGPT to write an email, or watched an AI turn a one-line prompt into a picture, you’ve already used generative AI. Most people have. What most people don’t have is a clear idea of what the term actually means.
This guide answers the question in plain English. What is generative AI, how does it work, what can it create, and what should you watch out for? No jargon, no hype, just the parts worth knowing.
What is generative AI?
Generative AI is artificial intelligence that creates original content in response to a request you type (or say). Give it a prompt and it can produce text, images, code, audio, or video that didn’t exist before.
The “generative” part is the key word. Older AI systems mostly recognised things or made predictions. This new wave generates things. That single difference is why AI suddenly feels so visible in daily life.
If you want the wider picture first, our beginner explainer on what AI is covers the basics in a few minutes.
How is it different from the AI we already had?
AI has been working quietly in the background for years. Your spam filter decides which emails look suspicious. Netflix predicts what you might watch next. Your bank flags a card payment that doesn’t fit your pattern.
All of that is traditional AI. It sorts, ranks, and predicts. It never writes you a poem.
Generative AI flips the job around. Instead of labelling content that already exists, it produces new content on demand. Same underlying family of technology, very different output. A useful shorthand: traditional AI answers “which one?”, generative AI answers “make me one”.
How does generative AI actually work?
The short version: these systems are trained on enormous amounts of text, images, and other data. During training, a neural network learns the patterns in that data, which words tend to follow other words, what edges and shapes make up a cat photo, how code is usually structured.
The result of all that training is called a foundation model. When you type a prompt, the model uses everything it learned to predict a fitting response, one small piece at a time. Chatbots like ChatGPT are built on a specific type called a large language model, which does this with text.
It’s not magic and it’s not thinking. It’s extremely good pattern prediction at a scale no human could match. AWS has a solid technical explainer if you want to go one level deeper.
What can generative AI create?
- Text: emails, summaries, study notes, articles, translations. Tools: ChatGPT, Gemini, Claude.
- Images: illustrations, product mockups, social graphics. See our guide to AI image generators for beginners.
- Code: working snippets, bug explanations, whole small apps.
- Audio and video: voiceovers, music drafts, short generated clips.
From my own work on websites and digital projects, text and code are where beginners get value fastest. Drafting a page, summarising a long document, or explaining an error message takes seconds instead of an hour.
Why did this suddenly become such a big deal?
Researchers worked on generative models for years, but the turning point for the public was ChatGPT’s launch in late 2022. For the first time, anyone could type a plain sentence and get useful output back, no technical skills needed. Since then, generative features have been built into search engines, email apps, office software, and phones.
In other words, you no longer go to generative AI. It comes to you.
The limits you should know about
Generative AI predicts what a good answer looks like. It doesn’t check whether that answer is true. Sometimes it produces confident, wrong information, a problem covered in our guide to AI hallucinations. It can also repeat biases from its training data, and questions about copyright on generated content are still being settled.
Working around cybersecurity, I’d add one more habit: don’t paste passwords, client data, or anything sensitive into an AI chat. Treat it like a public place, not a private notebook.
Tip: use generative AI for first drafts and explanations, and keep yourself as the final editor. Verify any fact, name, or number before you rely on it.
How to try it yourself (free)
You don’t need to install anything. Open a free chatbot and give it a real task from your day: “rewrite this email so it’s shorter and friendlier” or “explain this paragraph like I’m 12”. You’ll learn more from ten minutes of doing than from any definition.
If you prefer something structured, Google’s free Introduction to Generative AI course takes about 45 minutes and assumes no background.
Common Questions
Is generative AI the same thing as ChatGPT?
No. ChatGPT is one product built on generative AI. Gemini, Claude, and image tools like those in Canva are others. Generative AI is the category, not the app.
Do I need technical skills to use it?
No. If you can describe what you want in a normal sentence, you can use generative AI. Clearer requests get better results, but there’s nothing to code or configure.
Can I trust what generative AI tells me?
Mostly, but not blindly. It’s reliable for drafting, summarising, and explaining, and less reliable for facts, figures, and anything recent. Double-check important claims against a trusted source.
Final takeaway
Generative AI is simply AI that creates: text, images, code, and more, from a plain request. It’s a prediction machine, not a truth machine, so use it to work faster and keep your own judgement in charge. Start with one small daily task this week, and it will earn its place in your routine quickly.
