One other week introduced one other random e mail from a fan. They puzzled why my writing is “so completely different” between this weblog — my private weblog — and writing they’ll discover elsewhere. Properly, I’ve no ethical, moral, or authorized requirement to elucidate myself to anybody on the subject of my writing. What I do in my non-public time (normally from the consolation of my dwelling workplace or the commuter practice) is my enterprise.
Positive, in the event you’re paying me to write down, then you have got some (notice: some) proper to know the way I do my work, and why I selected to write down a few explicit topic. Should you’re “tipping” me for my writing through some form of subscription or precise tipping, there may be some duty on my half in explaining to you ways and why I wrote what I wrote. For instance, on my Medium.com account, in the event you’re a Medium member and have questions on my writing course of, I may be extra inspired to reply. Should you’re an off-the-cuff reader and wish to know extra… Properly, I’d take into account the request and act on it (or not).
However on this weblog? Nah, man. That is my weblog. I pay WordPress to keep up the positioning and assist me write my ideas and put them out into the world. I don’t promote, so there isn’t a management from any third social gathering in any respect. It’s all me, child.
However why is it “so completely different”? The reply is evident. That is my private weblog, so I normally write it multi functional go. Infrequently, I’ll get interrupted and end in a number of classes. However I hardly do any enhancing (aside from the grammatical and orthographical corrections from Grammarly, which is what I exploit to proofread most of my writing). This weblog is nearly (nearly) stream of consciousness.
After I write on Historical past of Vaccines, a extra delicate course of is concerned. I’ve to discover a good matter, normally one that’s not being totally addressed some other place. Then I do background analysis. Then I write a primary draft. I then work on it for a number of days, then run it by one or two specialists if the subject material is one thing I’m not an skilled in myself. After which it will get printed.
After I wrote on Medium, the method is much like writing on Historical past of Vaccines. However Medium shouldn’t be my full-time job, so the time scale is longer, and normally entails researching and writing late into the night, particularly after the child goes to mattress. Most of my articles recently have been printed through publications on Medium, which have their very own editors. These editors should agree, after which the article will get printed. And, as I mentioned above, members get precedence in responses.
After which there’s the Substack Public Well being Night time College venture. As with Medium, people who find themselves paying for a subscription get first dibs at feedback and discussions of the matters. Additionally as with Medium, the method of writing there is identical. It’s simply simpler as a result of I’ve extra experience on Public Well being and its subtopics, so writing these weblog posts comes simple.
What concerning the 500-pound gorilla within the room on the subject of creativity — particularly round writing — recently? The place do Giant Language Fashions (aka Synthetic Intelligence) match into my workflow? Properly, I exploit Perplexity.ai to do my analysis. Then I exploit Grammarly to repair grammar, punctuation, and spelling errors. After which the next immediate on ChatGPT helps me discover holes in my writing:
Please learn the next draft I’ve written and clarify it again to me like I’m a highschool scholar. If there are any “holes” within the narrative, level them out to me. Ask me questions if one thing shouldn’t be clear so I can higher tailor the draft to elucidate the whole lot I’ve written. Don’t change something within the draft. Don’t add/delete from it. Don’t make something up by sticking to the topic at hand within the draft.
Oh, sure, I additionally use ChatGPT to generate photos, and to assist me code for information evaluation initiatives. It is usually a superb instrument to assist me craft my lecture shows in methods which can be extra comprehensible, utilizing the same course of to what I point out above. Due to the excessive availability of those Giant Language Fashions, I don’t assign many writing assignments anymore. And I’m skeptical of scholars who get good grades on quizzes.
However right here’s my take… I train graduate college students, not kids. I depart it as much as them whether or not or not they’re going to cheat. In any case, they’re those accountable to be taught. In the event that they don’t be taught due to dishonest, that’s on them. It’s going to return up later of their careers. When the time comes to use the data in actual life, it’s going to be on them.
I simply don’t wish to waste my time grading one thing that has a excessive likelihood of being created by AI.
So, yeah, there will probably be little bits of AI work right here and there in my work. In relation to what I’m getting paid for, I’ve an aversion to dishonest. So I don’t do it. Or, relatively, I strive to not do it. There are a lot of occasions when Grammarly’s or ChatGPT’s recommendations for writing are higher than my writing, particularly on the subject of passive writing. So I exploit them as instruments to boost what I write, to not do the writing itself. And, sure, I every now and then use these instruments to do some brainstorming.
Anyway, no, I don’t owe you a proof, despite the fact that I simply gave you one.
Thanks for studying.
