Saturday, April 08, 2023

8: ChatGPT

I use these 5 ChatGPT hacks to cut my workload by 8 hours a week. Here's the exact wording of my prompts. I've been in marketing for 18 years. ChatGPT has helped me make my workflow more efficient. ....... The generative AI helps me write social-media posts, troubleshoot marketing problems, and more. ...... In my 18-year marketing career, I've written millions of words of marketing copy. ....... I'm now a chief brand officer for a real-estate-coaching business. Language-modeling AI tools like ChatGPT quickly get me 80% or 90% of the way to finished work. ........ I use it for dozens of things every day, but these five hacks have been the most helpful. ....... I prompt ChatGPT with: "Write me a [number]-word social media post/email in a conversational tone like Jess Lenouvel using the following outline," adding three or four bullet points. ...... Adding our CEO's name to the prompt helps ChatGPT mimic her writing tone, cadence, and even emoji usage. ...... "Think like a digital marketer and generate four variants of this Facebook ad copy." ...... Check and modify the copy's tone and reading level ........ People want easily understandable content in their inboxes and on social media. The average American is thought to read at a seventh- or eighth-grade level. ........ I think the best marketing copy reads a little below that, at a fifth- or sixth-grade level. This means short sentences and paragraphs, a conversational tone, and simple language. ....... Before ChatGPT, I'd use Hemingway to test the reading level of my copy and fiddle with sentences to hit a fifth-grade level. With ChatGPT, I can write freely and then prompt the AI to change the reading level and tone for me........ Write long-form storytelling posts by giving the AI a specific structure and outline ....... Storytelling is the backbone of psychology-based marketing. ........ "Following this story structure — 1. Capture the heart, 2. Set up a tension, 3. Resolve the tension, 4. Conclude by offering value — write a 1,000-word story at a grade-five reading level in the first person using the following information," followed by a list of plot points. ........... "Write a 400-word social media post in a conversational first-person tone like Jess Lenouvel about [topic] using the following story points." ......... "I'm running an A/B test to compare application page A and application page B in order to increase the number of applications to our program, and I need help generating hypotheses based on email click rate and form completion. Can you provide recommendations for what to test and how to measure success?" ........... I interact with other marketers on Discord servers and in Facebook groups to get more ideas for using ChatGPT. ........ On average, ChatGPT saves me between six and eight hours a week......... Almost every time I've asked myself "I wonder if ChatGPT can do this" or "I wonder how I could get ChatGPT to help me with this," I've unlocked a new way of streamlining or optimizing my workflow. .

How AI experts are using GPT-4 Plus: Chinese tech giant Baidu just released its answer to ChatGPT. ......... Microsoft wants you to use GPT-4 in its Office suite to summarize documents and help with PowerPoint presentations—just as we predicted in January, which already seems like eons ago. ........ Google announced it will embed similar AI tech in its office products, including Google Docs and Gmail. That will help people draft emails, proofread texts, and generate images for presentations. ......... Narayanan says he’s been testing AI tools for text generation, image generation, and code generation, and that he finds code generation to be the most useful application. ........ Baidu unveiled a new large language model called Ernie Bot, which can solve math questions, write marketing copy, answer questions about Chinese literature, and generate multimedia responses. ......... performs particularly well on tasks specific to Chinese culture, like explaining a historical fact or writing a traditional poem .......... Large language models are infamous for spewing toxic biases, thanks to the reams of awful human-produced content they get trained on. But if the models are large enough, they may be able to self-correct for some of these biases. Remarkably, all we might have to do is ask. .

Battle of the chatbots: how does GPT-4 stack up against Bard? OpenAI and Google have been opaque about how their models were built. However, it is likely their training data and objectives are distinct. ....... GPT-4 is also disconnected from the internet and only has knowledge of events until September 2021. Bard can ostensibly bring results from Google search, although that does not seem to enhance the quality of its responses....... At present, I am not worried for the careers of the world’s diplomats. .

No comments: