I’m hosting a Professional Development session about Responsible AI policies in the classroom on 29th April 2026, and would love it if you could join me in a conversation about how to navigate the evolving AI landscape. We’ll briefly cover AI detection, what other educators are doing, and media literacy in the age of AI. I’ve recently become a GPTZero Ambassador (for those who don’t know, GPTZero is an AI writing detector), so all attendees of this webinar will receive 1 semester of GPTZero Premium for free! Join here: 29 April 2026, 10 am BST https://meet.google.com/qnd-kdtg-epf

(Steps 4 and 5 are the ones nobody teaches.) Most researchers treat "major revisions" as a near-rejection. It isn't. It is an invitation to publish, with conditions. Here are the rules I use, and now teach my PhDs and postdocs. 1) Triage before you feel Open the decision letter and make three columns: factual errors to fix, reframing the reviewer wants, and objections you will push back on. No emotion on this pass. 2) Write a one-line summary of each reviewer "R1: wants a broader intro, sceptical of the sample size. R2: accepts the contribution, wants tighter results. R3: silent on methods, has issues with our writing." Knowing who each reviewer is before you write the response saves you from arguing with the wrong one. 3) Answer in their order, not yours Reviewers reread their own comments. If your response letter jumps around, they scan for their point and get annoyed when they can't find it. Mirror the structure of the original report. 4) Quote, then respond Paste each reviewer comment verbatim in a different colour or italics. Respond underneath. As journals are short of good reviewers, such an approach will lower their cognitive load, and they can approve your work in a single pass. 5) Thank, then push back When you disagree, open with one sentence that names the merit of the reviewer's point, and then lay out your counter-evidence. Never use flat "no". Use a specific "I hear the concern, and here is why the data point the other way". One paper. Cleaner response letter. Faster turnaround. Save this for your next major revisions letter.

If you want to have a look at how Octopus actually looks like, here's the link to my two recently published minute publications. https://www.octopus.ac/publications/yxks-sg07/versions/latest It is a piece of work I've been doing as part of the UKCCSRC project, and I am currently working on the full paper. I thought, let's use it as a template or example for the research community to kind of explore how the entire process works with micropublications. And I just want to test whether the journals will be keen on actually publishing my paper. So I'll be updating everyone as we go along. Here's the link. You can see it does have a DUI number. It links the two papers so you can clearly see how the output is shaping up. Let me know whether that is helpful.

Yesterday I was interviewed by the Korean Broadcasting System (the BBC equivalent) about Teesside's transition to net zero. [This wouldn't be possible if I only published papers!] If you want to accelerate your academic career, try doing these in addition to publishing your research. 1. Seek external engagement: Publishing academic papers is vital, but stepping outside the academic circles will amplify your impact. Collaborating with media outlets and professional bodies can help showcase your expertise and research to a broader audience. 2. Build an expert brand: Visibility is key. Share your insights on various platforms (i.e. conferences, podcasts, or documentaries). This not only enhances your reputation but also helps position you as a thought leader in your field. 3. Engage with the community: Connect with peers and practitioners outside your usual circles. Networking and sharing your research with those who can benefit from it fosters innovation and collaboration. I apply ALL of this to my life and there is not a single day I am not grateful for keeping my focus on the things that are most important. How do you plan to take your research beyond the academic circles?

Ever since we started using computers, we have been taught to type everything on keyboards. And as a result, this is how we interact with the machines. So whether you write papers, you write responses to emails, you write LinkedIn posts, anything, essentially, you're just typing. Recently, I was introduced to VoiceInk, an application that lets you dictate what you want to say and transcribes it into text. I know I've been trying to use Google Docs for this. I've been trying to use MS Word for this and then including some of the transcription services available on Mac. But I always found that it was fairly inaccurate. I couldn't really rely on it because the grammar was off, and it missed words because of my non-native accent. But now we see that AI tools actually make it much more profound. So you can essentially leverage AI to understand how you speak and convert your speaking to text with very high accuracy. So the tool that I'm using, as I mentioned, Voice Inc, it uses some of its own AI tools, but you can also hook it up to OpenAI or any other tools that you have available. You can use open source tools like Ollama. To get that AI enhancement. And it has already saved me quite a lot of time. So I started using that two weeks ago. Right now, it has saved me about 40 minutes to 60 minutes of my time. It's a little bit weird for me to speak to the PC rather than type. So I'm still learning. But I thought I would share this with you. I know we are all kind of struggling with time, and essentially, you speak faster than you write. So that could actually help you with writing papers because you can just start talking about your paper, talking about your results, talking about your discussion, and then just fix the grammar, fix the writing and everything else. So that's one of the main reasons why we have a writer's blog, because I posted on LinkedIn today. We do want to make sure our writing is perfect. And as a result, we write and then we edit, we write and then we edit. And as a result, we waste time.