Last December I experimented a bit with the then new at the time ChatGPT writing haiku. It turns out that particular post has been quite popular – among the most visited since I shared it.
This morning I set out to use Google Bard (now powered by the Gemini Pro model) to write haiku using the same prompts as last year with ChatGPT. Here are the results – which I also find very enjoyable!
Note: The image above was created using Bing’s Image Creator powered by OpenAI’s DALL-E 3.
Write a haiku about film photography
Sun paints on silver, A patient click, time captured still, Light’s memory held.
Write a haiku about a quiet morning in a boat on a lake
Mist hugs sleeping reeds, Oars dip, a whisper in glass, Sun’s kiss wakes the lake.
Write a haiku about snow falling on cedars
Jade boughs bow with grace, Each flake a silent, feathered kiss, Whispers on the breeze.
Write a haiku about stillness in Kyoto
Temple bells toll slow, Bamboo sighs in ancient shade, Stone gardens breathe deep.
Write a haiku about watching the first rain of the year
Parched earth drinks the first sip, Rooftop whispers secrets soft, Green whispers unfurl.
Write a haiku about geese flying overhead
With beating wings that cleave the sky, A honking skein in chevroned flight, Winter’s wild grace calls.
Write a haiku about farming with a John Deere tractor
Green giant hums low, Iron claws till earth’s slumbering heart, Seeds dream of sunlit rows.
Write a haiku about writing haiku
Few words paint a world, Seventeen whispers hold the breath, A flower unfolds.
And now for a few more…
Write a haiku about visiting Stockholm in late December
Write a haiku about walking across the Golden Gate Bridge in fog
Towers pierce white mist, Footsteps echo, lost in air, City sleeps unseen.
Write a haiku about the joys of Christmas with family and friends
Laughter fills the hearth, Gifts unwrapped, eyes shining bright, Love warms winter’s breath.
Write five haiku about the joys of walking a dog in the park on a cool grey morning
Crisp air nips at cheeks, Tail wags a metronome beat, Leaves crunch under paws.
Squirrels taunt from oaks, Bark explodes in joyous chase, Frisbee paints the sky.
Dew kissed blades stand tall, Golden fur a furry sunbeam, Sniffs of morning dew.
Pigeons flutter up, Startled by a playful lunge, Laughter fills the park.
Back to leash, head home, Whiskers twitching, ears perked high, Memories in paws.
And then I asked Google Bard for some judgment:
Which one is the best?
I think the fourth haiku is the best! It captures the playful energy of a dog walk in the park, as well as the delightful surprise of the startled pigeons and the shared laughter between dog and owner.
Here’s why I think it stands out:
Vivid imagery: The line “Pigeons flutter up” creates a clear picture in the reader’s mind, and the “Startled by a playful lunge” adds a touch of humor.
Sound: The “Laughter fills the park” adds another sensory element to the haiku, making it feel more immersive.
Connection: The laughter shared between dog and owner highlights the joy and companionship that comes from spending time with our furry friends.
Overall, I think the fourth haiku is the most evocative and memorable of the bunch. It’s a great example of how haiku can capture a moment in time and leave a lasting impression on the reader.
Of course, the best haiku is ultimately a matter of personal preference. Some people might prefer the serene beauty of the first haiku, or the joyful chase of the second. But I believe the fourth haiku stands out for its combination of vivid imagery, sound, and connection, making it a truly delightful read.
Remember that mind-blowing Google I/O demo of an AI tool that unlocks hidden insights from your research documents? That’s NotebookLM, and it’s not just for tech giants anymore. (See this earlier blog post about what was originally Project Tailwind.)
As a longtime reader of author Steven Johnson (and avid follower of his “Adjacent Possible” Substack), I was thrilled to learn he’s now part of the team at Google Labs bringing this powerful technology to the masses.
Imagine uploading piles of research papers, articles, or even future forecasts (like I did with those year-end reports from Wall Street investment houses forecasting what’s expected in 2024!), and then having NotebookLM not only summarize them but also weave connections you might have missed. That’s exactly what I experienced.
NotebookLM’s “additional questions” feature is a game-changer, prompting me to explore angles I wouldn’t have considered on my own. It’s like having a tireless research assistant with an uncanny knack for spotting crucial details.
Of course, NotebookLM is still in its early stages. The current 20-document limit can feel restrictive, and its future as a paid product is unclear. But for researchers grappling with mountains of information, it’s a game-changer. It’s not just about saving time; it’s about sparking genuine intellectual leaps.
This tool isn’t just for academics, though. Imagine journalists using NotebookLM to connect seemingly disparate news articles, or students piecing together complex historical narratives. The possibilities are endless.
Sure, like any AI tool, it’s not perfect. Fact-checking is crucial, and occasional “hallucinations” can crop up. But NotebookLM’s source citations make verification easier, and its overall accuracy is impressive so far.
So, ditch the highlighter and embrace the future! NotebookLM isn’t just a fancy research tool; it’s a bridge to deeper understanding, more insightful analysis, and ultimately, groundbreaking discoveries. Unleash your research potential – your next breakthrough might just be a question away.
Note: This morning I read an article in the Washington Post about Charles Bailey who has been nominated to lead Montgomery County Maryland’s Fire and Rescue Service. In the article, Bailey is quoted: “You look at a piece of art that you know nothing about, and you ask yourself three questions: Function, meaning, style. And if you can answer, or derive answers to those three questions, it allows you to place the work of art into a context, which is the first step in deriving some sort of meaning from it.” I liked his comments and asked Google Bard to help flesh out these ideas.
Here I stand, face to face with an enigma. A canvas, a photograph, a silent symphony of pigment and light. I know nothing of its history, its creator, its purpose. Yet, a spark ignites within me, a hunger to understand. And so, I turn to my three trusty companions: Function, Meaning, Style.
Function: What purpose does this piece serve? Is it a portrait meant to capture a soul, a landscape yearning to share its grandeur, or perhaps a social commentary holding a mirror to society’s flaws? The brushstrokes, the composition, the very subject matter – each whisper hints at the artist’s intent. Is it a testament to faith, a celebration of beauty, or a cry for change?
Meaning: Beyond the surface, what whispers beneath the paint? What story does it yearn to tell, what emotions does it evoke? Is it a lullaby of serenity, a war cry of defiance, or a bittersweet ode to lost love? The interplay of light and shadow, the tension between form and space, each element becomes a brushstroke in the narrative. Do I see hope or despair, joy or sorrow, a call to action or a melancholic sigh?
Style: The artist’s voice, their unique language, speaks through the chosen medium. Is it the bold, expressive brushwork of a Van Gogh, the meticulous realism of a Vermeer, or the abstract whispers of a Kandinsky? Each stroke, each shade, reveals the artist’s soul, their relationship with the world. Do I see rebellion or reverence, chaos or order, a whispered secret or a deafening roar?
As I delve deeper into these questions, the canvas begins to come alive. The brushstrokes dance, the colors sing, the subject reveals itself. The enigma takes shape, not as a single, definitive answer, but as a kaleidoscope of possibilities. This is the beauty of art – it invites us to co-create, to fill in the blanks with our own experiences, emotions, and interpretations.
And so, I stand before the enigma no longer a stranger, but a fellow traveler on the path of meaning. The answers may remain elusive, but the journey itself is a revelation. For in the act of questioning, of seeking, I discover not just the meaning of the art, but also the echo of my own humanity within its depths.
This, I believe, is the true function of art – to not merely adorn walls, but to illuminate the labyrinthine pathways of our own existence. And as I turn away from the canvas, carrying its whispers within me, I know that the enigma has not been solved, but rather, transformed. It has become a mirror, reflecting not just the artist’s intent, but the ever-evolving story of who I am.
Interacting with an AI assistant like Claude is a two-way conversation. To get the most out of your digital assistant, you need to know how to provide useful prompts that set it up for success. Follow these prompt writing tips when chatting with your AI:
Be Direct and Clear
Don’t beat around the bush when posing questions or requests to your AI assistant. Vague or ambiguous prompts will lead to vague and ambiguous responses from the AI. Get straight to the point and be as specific as possible about what you want it to do. Break down complex instructions into simple, step-by-step prompts to guide the AI.
Provide Sufficient Background
Don’t assume your AI assistant has any implied common sense or knowledge about your request. The AI only knows what’s in its training data! Give sufficient background and context in your prompts to set the AI up to respond appropriately. If you’re asking about a specific person or event the AI isn’t likely familiar with, give a brief explanation.
Ask Focused Questions
Open-ended questions can sometimes lead your chatbot down useless tangents. If you want specific info, ask specific follow-up questions rather than overly broad ones. You’ll get better results by being laser-focused on the information you need.
Check Responses and Follow Up
Your AI assistant is learning all the time. Check its responses to see if they actually provide what you were looking for, and follow up if it seems off track. You can rephrase your original prompt or ask clarifying questions to get the AI back on a useful path. Consider it a collaborative process.
Test the Limits
Feel free to get creative in testing the limits of your AI assistant’s capabilities. Novel prompts encourage the AI to expand what it can handle. Just be sure to provide plenty of guidance and background if you’re requesting something unusual or complex.
Give Your AI a Role to Play
When chatting with your AI assistant, it can be hugely beneficial to explicitly give it a role to play, such as teacher, travel agent, cooking assistant, or even friend. Defining a role provides necessary context that your AI needs to adapt its responses and language appropriately.
Without any framing, your assistant has no way of knowing how to act or what kind of information you need. A vague conversation with an undefined AI assistant can quickly go off the rails. But when you assign a clear role, it focuses the dialogue and reduces misunderstandings.
The role you give your AI doesn’t have to be rigid or limiting. Think of it as providing a general framing, tone, and personality that fits your needs for that particular conversation. Just tell your assistant “Act as my tutor and explain this concept to me.” Or “Pretend you’re a museum tour guide and describe this painting.”
Taking on a role allows your AI to access the right knowledge, terminology, and communication style to have a productive, natural dialogue. With the power of role-playing, your AI assistant can become an even more effective conversation partner!
Some Prompt Examples You Can Try
To see your AI assistant’s skills in action, give some longer, more descriptive prompts like these a try:
“Imagine you are an art museum tour guide named Sam leading a group through the impressionist gallery. In an enthusiastic but professional tone, explain the key characteristics of Monet’s painting style and describe what makes his Water Lilies painting a quintessential example of impressionism.”
“Act as a friendly chess tutor named Alex teaching me, a novice player, strategies for the first 10 moves as white in a classic game opening. Use simple, straightforward language to clearly explain the objectives of different standard openings and why they are solid foundational openings for white.”
“Pretend you are a home repair specialist named Jamie and I’m a customer needing help fixing my broken kitchen faucet. Walk me step-by-step through how to remove the faucet handle, disassemble and inspect the internal parts, replace any faulty washers or O-rings, and reassemble everything to get the faucet working again.”
“You’re an accomplished pastry chef named Pat teaching me how to bake chocolate chip cookies from scratch. Explain in a warm, patient tone the key steps from mixing dry and wet ingredients properly to getting the baking time and temperature right. Share any tips to ensure the cookies turn out perfectly chewy.”
“Imagine you are a professional genealogist named Gene helping me trace my ancestry. I’ve hit a roadblock finding records about my great-grandfather who I believe was born in Ireland around 1885 before immigrating to New York City. In a warm, encouraging tone provide step-by-step guidance for how to locate passenger manifests, naturalization records, birth/marriage/death certificates, and other documents that could shed light on his origins and immigration journey. Feel free to ask clarifying questions if you need any additional details from me. Your role is to patiently coach me through breaking through this genealogy brick wall using targeted record searches.”
Prompt for Haiku
To showcase your AI’s creative writing skills, try a prompt like:
“Write a haiku poem about viewing cherry blossoms in spring. Describe the delicate pink flowers blooming on branches and scattering in the wind. Convey a tranquil, reflective mood. Follow the 5-7-5 haiku structure with 3 lines and appropriate season word.”
This provides context about the haiku’s theme, asks the AI to use descriptive language and imagery, gives guidance on mood and structure, and defines the specific creative writing form. Framing poetic prompts in this way allows your AI to tap into its artistic expression. You can guide it to compose haiku, limericks, sonnets, and more on any topic you choose!
Giving longer prompts with more conversational detail and specific instructions allows your AI assistant to generate richer, more natural responses. Don’t be afraid to really set the scene and adopt a persona when framing your requests! For even more, see my post Multi-step prompting for Claude.
Please share what you’ve learned about effective prompts that you’ve discovered!
Credit: Drafted with the assistance of Claude, an AI assistant created by Anthropic.
Abstract surface of Multicolored splash watercolor blot. Artistic hand-painted vector, element for banner, poster, card, cover, brochure.
On my walk this morning I listened to the latest episode of Cal Newport’s podcast Deep Questions which was about “creativity hacking”. He reviewed a couple of his techniques, shared a few links to good articles related to the topic, etc. For example, he talked about using different venues for working – such as moving between multiple locations in a single day. He cited several other techniques that he’s used all of which involved helping lift your mind out of the mundane to stimulate its ability to get creative.
Listening to Cal stimulated my mind to make the bridge between something like working on a whiteboard in front of a group of colleagues and, in lieu of humans, working on a topic with an AI chatbot like Claude, Bard, ChatGPT, etc.
I’m drawn to this approach because of just how low the overhead is to start to use chatbots for helping in my thinking and creativity processes. Unlike other humans, the chatbots are always available – 24×7 – and generally they’re pleasant and polite to interact with. Sometimes humans are too but they can also be unpleasant and impolite!
The speed of the interaction with a chatbot is probably slower than with another human or group of humans but I don’t find that to be an issue. In some ways, the ability to just pause, give me time to think, and then interact again with a chatbot is more pleasant and completely eliminates any peer pressure I might be feeling.
I can also use chatbots to respond to challenging questions I might ask or have it brainstorm wild ideas, or roleplay different perspectives – all of which help to get my creative juices flowing. A friend would do the same – but, again, the low overhead nature of creativity hacking with a chatbot is very attractive to me. I can also spread my chatbot interactions around between the various AI chatbots. They’ll bring different perspectives, just like a group of my friends might. I might also ask a chatbot to help me think about an idea from the perspective of a particular historical figure (e.g., “what would Steve Jobs say about this idea?”).
I’m increasingly finding the benefits of this chatbot approach helping me in my creative work. For example, while writing this blog post, I asked for a critique of it from one of the chatbots. Over a few interactions, I was able to gain some new insights to include which hopefully made it even better.
Update: this morning Andrew Chen posted a great article titled “How I use AI when blogging and writing” which explores his experiences using AI chatbots to help with blog posts, generating topic ideas/questions, and even starting a book outline. Definitely worth reading!
I was intrigued by this sentence in a recent column by Bloomberg’s Matt Levine: The stereotypical way that a management consulting interview works is that you get asked questions like “how much pepperoni is sold in the US every year,” and then you reason your way to some estimate of the answer.
When I read this, I thought it would be fun to play with my favorite chatbots again and see if they might help me reason my way to an answer. I got very different results – not surprisingly. The prompt I used was:
The stereotypical way that a management consulting interview works is that you get asked questions like “how much pepperoni is sold in the US every year,” and then you reason your way to some estimate of the answer. Let’s reason step by step to try to find an answer to this question.
First, here’s what I got back from Google Bard:
Sure, here is a step-by-step process of how to reason our way to an estimate of how much pepperoni is sold in the US every year:
Gather some basic facts. We know that the average American eats 24 pounds of pizza per year, and that pepperoni is the most popular pizza topping, accounting for about 30% of all pizza orders. We also know that there are about 330 million people in the US.
Make some assumptions. We need to make some assumptions about how much pepperoni is used per pizza. A conservative estimate might be that each pizza uses 1 ounce of pepperoni. This means that the average American eats about 2.4 ounces of pepperoni per year.
Do some math. We can now estimate that the total amount of pepperoni sold in the US every year is about 800 million ounces. This is equivalent to about 250 million pounds.
This is just an estimate, of course, but it is based on some reasonable assumptions. It is likely that the actual amount of pepperoni sold in the US is higher than this, but it is also possible that it is lower. More accurate estimates could be made by collecting more data and making more refined assumptions.
Here are some additional factors that could be considered to improve the accuracy of the estimate:
The amount of pepperoni used per pizza can vary depending on the size of the pizza, the type of pizza, and the restaurant.
The popularity of pepperoni may vary from region to region in the US.
The amount of pepperoni imported into the US may also be a factor.
By considering these factors, it is possible to make a more accurate estimate of the amount of pepperoni sold in the US every year.
Next up was ChatGPT:
Of course, I’d be glad to guide you through the process of reasoning to estimate the amount of pepperoni sold in the US every year. It’s important to note that management consulting interviews often focus on assessing your problem-solving skills and ability to think logically. Let’s break down the problem step by step:
Step 1: Clarify the Scope Before we begin, it’s essential to clarify the scope of the question. Are we looking for the total weight of pepperoni sold, or the total revenue generated from its sales? This distinction will influence the data sources we’ll need to consider.
Step 2: Identify Relevant Data Sources To estimate the amount of pepperoni sold in the US, we’ll need to gather data from reliable sources. These could include industry reports, government statistics, market research firms, and food industry associations. Some possible sources might be the U.S. Department of Agriculture (USDA) and organizations like the American Meat Institute.
Step 3: Estimate Total Consumption One approach could be to start with the total meat consumption in the US and then estimate what percentage of that is likely to be pepperoni. You might need to consider factors such as the popularity of pizza, sandwiches, and other dishes that use pepperoni. Additionally, think about variations in consumption due to cultural and regional preferences.
Step 4: Break Down Consumption Break down the total estimated consumption of pepperoni into categories such as retail sales (supermarkets and grocery stores) and food service (restaurants, pizzerias, etc.). This can help refine your estimate by considering different distribution channels.
Step 5: Estimate Quantity per Serving and Servings Estimate the average quantity of pepperoni used per serving. For example, how many slices of pepperoni are typically used on a pizza or in a sandwich? Then, multiply this by the estimated number of servings sold in a year.
Step 6: Adjust for Factors Consider any factors that might affect the estimate, such as seasonal variations in consumption (for instance, more pepperoni might be consumed during holidays or certain events) or any recent trends that might have impacted consumption patterns.
Step 7: Calculate the Estimate With all these factors in mind, you can proceed to calculate your estimate. Make sure to show your calculations and assumptions clearly.
Step 8: Validate and Compare After arriving at your estimate, it’s a good practice to validate it against any available data, if possible. Compare your estimate to industry reports or official statistics to assess its accuracy.
Remember, the goal of this exercise is not necessarily to arrive at an exact figure, but to showcase your ability to logically reason through a complex problem using available information. Feel free to ask questions if you need further clarification on any of these steps.
Next up is Claude:
Okay, let’s break this down:
First, we need to estimate how much pizza is sold in the US each year. Pizza is a very popular food, so let’s assume that on average, each American eats pizza once per week. With a population of around 330 million, that’s about 330 million pizzas sold per week, or around 17 billion pizzas per year.
Next, we need to estimate what percentage of those pizzas have pepperoni. Pepperoni is a very common pizza topping, so let’s assume it’s on 50% of all pizzas sold. That brings us to around 8.5 billion pepperoni pizzas per year.
Now we need to estimate how much pepperoni is on each pizza. Let’s assume an average pepperoni pizza has around 50 pepperoni slices. If each slice is roughly 1 ounce, that’s about 50 ounces or around 3 pounds of pepperoni per pizza.
With 8.5 billion pepperoni pizzas per year, each with around 3 pounds of pepperoni, that gives us an estimate of around 25.5 billion pounds of pepperoni sold annually in the US.
Of course this is just a rough estimate with a lot of assumptions. The actual amount may vary quite a bit. But walking through the logical steps and making reasonable assumptions is generally the approach for these types of brainteaser questions. Let me know if this helps explain the thought process!
Finally is Bing Chat:
That’s an interesting question. To estimate how much pepperoni is sold in the US every year, we can use some assumptions and data from the web search results. Here are some possible steps to follow:
First, we can assume that most of the pepperoni sold in the US is used as a pizza topping, since it is one of the most popular pizza toppings.
Second, we can use the data from 7 to estimate how much pepperoni is used on a pizza of different sizes. For example, a 12-inch pizza uses about 1.75 ounces of pepperoni, and a 18-inch pizza uses about 3 ounces of pepperoni.
Third, we can use the data from 6 to estimate how many pizzas are sold in the US each year. According to this source, there are 3 billion pizzas sold in the US each year.
Fourth, we can use the data from 13 to estimate what percentage of pizzas ordered are pepperoni. According to this source, 36% of people order pepperoni on their pizzas.
Fifth, we can multiply the number of pizzas sold by the percentage of pepperoni pizzas, and then multiply by the average amount of pepperoni per pizza. To simplify the calculation, we can assume that the average pizza size is 14 inches, which uses about 2.5 ounces of pepperoni according to 7.
Therefore, the estimated amount of pepperoni sold in the US each year is:
3 billion x 0.36 x 2.5 = 2.7 billion ounces
To convert this to pounds, we can divide by 16, since there are 16 ounces in a pound. This gives us:
2.7 billion / 16 = 168.75 million pounds
This is our final answer, based on the assumptions and data we used. Of course, this is only an approximation, and there may be other factors that affect the actual amount of pepperoni sold in the US each year.
I hope this helps you understand how to approach this kind of question. Do you have any other questions for me? 😊
Bard estimated 250 million pounds. Claude estimated 25.5 million pounds. Bing estimated 168.75 million pounds. ChatGPT didn’t provide an estimate – it just discussed a methodology to get to an answer (which is what’s more important in a job interview!).
What’s the correct answer? Some quick Google searching finds results around 250 million pounds a year – closest to Bard’s answer.
What can we learn from this little exercise? I would have expected both Bard and Bing Chat to have an advantage as they’re both capable of doing web searches while ChatGPT (3.5) and Claude aren’t. Bing actually cites specific web references. But the most useful result actually isn’t the final value – all of the chatbots demonstrated steps to go through to get to an answer and it’s the ability to demonstrate those steps that answers the interview question. I’m sure by continuing each of the conversations I could have gotten more insights from each of the chatbots. Meanwhile, I’m now hungry for a pepperoni pizza!
Update 04/21/2024: a few days ago Meta released its new Llama 3 model so I tried my original prompt out on the Meta.ai website. Here’s what I got back:
The classic “pepperoni problem”! Let’s break it down step by step. Here’s a suggested approach:
Step 1: Define the scope Clarify what’s meant by “pepperoni sold in the US every year.” Are we including:
Only pepperoni slices sold as a standalone product?
Pepperoni used as an ingredient in other products (e.g., pizzas, sandwiches, salads)?
Both domestic consumption and exports?
For this example, let’s assume we’re focusing on domestic consumption of pepperoni as a standalone product or as an ingredient in other food products.
Step 2: Identify key drivers What factors influence pepperoni sales in the US?
Population growth and demographics
Pizza consumption habits (as pepperoni is a popular pizza topping)
Trends in food preferences and dietary habits
Competition from alternative meat products
Step 3: Gather data points Collect relevant data to inform our estimate:
US population: approximately 333 million (as of 2024)
Pizza consumption: average American consumes 23 pounds of pizza per year (according to a 2020 survey)
Pepperoni’s market share among pizza toppings: assume around 30-40% (as it’s a popular topping)
Step 4: Make assumptions and calculations Based on the data points, make some assumptions and calculations:
Assume an average pepperoni slice weight: 2-3 grams ( rough estimate)
Estimate the total number of pepperoni slices consumed per year:
333 million people x 23 pounds of pizza per year ÷ 2-3 grams per pepperoni slice = approximately 12-15 billion pepperoni slices per year
Convert this to a weight-based estimate:
12-15 billion slices x 2.5 grams (avg. slice weight) = 30-37.5 billion grams or 30,000-37,500 metric tons of pepperoni per year
Step 5: Refine and validate Refine our estimate by considering additional factors, such as:
Seasonal fluctuations in pepperoni demand (e.g., summer BBQs, holiday parties)
Regional differences in pepperoni consumption
Market trends and competition
To validate our estimate, we could research industry reports, market studies, or sales data from major pepperoni manufacturers or distributors.
By following these steps, we’ve arrived at a rough estimate of 30,000-37,500 metric tons of pepperoni sold in the US every year. Keep in mind that this is a simplified example and actual numbers may vary based on more detailed analysis and data.
Recently I shared some techniques I’ve been accumulating for specific prompts that are useful when using various of the AI chatbots. Ethan Mollick, who inspired my earlier post, has just published a new article on his Substack: How to Use AI to Do Stuff: An Opinionated Guide.
In this comprehensive guide, Mollick offers valuable recommendations on selecting the appropriate AI chatbot tool for specific tasks. He covers a wide range of applications, including writing, creating images, brainstorming, making videos, working with documents and data, as well as acquiring information for learning purposes. I highly recommend giving it a read!
The stakes for this are quite high. People using AI have 30-80% higher productivity in some writing and coding tasks, and often feel happier having offloaded their most annoying work. That is a big incentive to learn to work with your intern. And this is really your intern.
If you’re learning about AI and interested in these kinds of things, follow Mollick on Twitter – he frequently shares tips and techniques based on his latest discoveries.
I happened to try something this morning – not sure what stimulated it – probably something on Twitter – anyway, I used the following prompt to ask Google Bard to summarize the tweets of a friend of mine. I entered:
Summarize the tweets of @nickmilo
Bard replied:
Nick Milo is a writer and thinker who tweets about a variety of topics, including note-taking, creativity, and personal development. His tweets are often insightful and thought-provoking, and he often shares links to interesting articles and resources. In recent tweets, he has shared a thread on the importance of idea emergence, a link to his newsletter on note-making, and a tweet about the importance of patience in the creative process.
Here are some of the key themes in Nick Milo’s tweets:
Note-taking: Milo is a passionate advocate for note-taking, and he often tweets about the benefits of taking notes. He believes that note-taking can help us to learn more effectively, to be more creative, and to live more mindful lives.
Creativity: Milo is also interested in creativity, and he often tweets about how to be more creative. He believes that creativity is a muscle that can be strengthened with practice, and he shares tips on how to come up with new ideas and to solve problems creatively.
Personal development: Milo is also interested in personal development, and he often tweets about how to live a more fulfilling life. He believes that we can all learn and grow, and he shares tips on how to set goals, to overcome challenges, and to live a more intentional life.
Overall, Nick Milo’s tweets are a valuable resource for anyone who is interested in learning more about note-taking, creativity, and personal development. His insights are thought-provoking and informative, and they can help us to live more fulfilling lives.
Here are some specific examples of Nick Milo’s tweets:
“Note-taking is a superpower. It’s how we learn, grow, and make sense of the world.”
“Creativity is a muscle. The more you use it, the stronger it gets.”
“Personal development is a journey, not a destination. It’s about learning and growing along the way.”
“If you want to be more creative, start by becoming a better observer. Pay attention to the details of your world.”
“Patience is an essential ingredient in the creative process. Don’t give up on your ideas too soon.”
If you are interested in learning more about note-taking, creativity, and personal development, I encourage you to follow Nick Milo on Twitter. His tweets are a valuable resource for anyone who wants to live a more fulfilling life.
Wow! I’m impressed with how well Bard described Nick’s tweets and his areas of focus in his creative work – as he’s described it in his tweets!
I’ve recently begun following Ethan Mollick – see his Twitter and Substack newsletter feeds. Ethan is very advanced in his skills for prompting AI chatbots to help him get some very interesting results. He’s also co-authored three recent papers on using AI in teaching which are available on SSRN.
Mollick was recently on the Possible podcast with Reid Hoffman and Aria Finger – it’s a great edition to listen to! I made some quick notes from a few of the suggestions he made in the conversation with Hoffman and Finger and from some of his tweets and newsletters:
“…the starting thing I would at least tell people to do that is the closest to a trick is to definitely give it context, tell it who it is and who you are. “I want to have a conversation with you as a… blank,” can really help. And then everything else kind of washes out because there’s so much subtlety in these conversations that we don’t know the answers to.”
“Cheap variation is very easy with AI. So, what I will do is say, “give me 40 versions of this paragraph in radically different styles,” and then skim through them for inspiration, right? “Give me 20 different analogies for this.” So I think it’s that power of tireless variation that I find super interesting.” (Scott’s note: I particularly like this use case because it avoids hallucination issues.)
“…it’s that inspiration piece — there was no way to do that before. I couldn’t ask an intern to do 20 different versions of a paragraph, right? There was no tool for that. So that, to me, is a little hack that actually has been pretty profound. Just do a lot of this, and then let me read a lot and figure out what the right answer is.”
“One technique I use often to get ChatGPT to give me the best evaluation of a topic is to ask it to steel man two sides of an argument, then write an opinion based on the two arguments.”
For Bing, “you should make sure you are forcing Bing to look something up with every query. Things that have worked for me include prompts like First research . Then do or else prompts like Look up _ on Reddit/in academic papers/in the news. Then use that to _. **Either way, you want to trigger the “searching for” label to get good results. The rules are still a little obscure as to what sorts of searches get triggered (does it look at specific URLs if you paste them in?) but experiment and you should be able to find something that works.”
“You can use this approach to focus Bing on a particular approach (Look up how Bain and Company does consulting analyses and then…), to learn new skills (a favorite of mine: Look up how to create image prompts using Midjourney and write a prompt that…) or to do more complex analysis.”
There are many more gems that he’s shared – so if you’re at all interested in learning how to be more effective using any of the AI chatbots, pay attention to what Ethan Mollick is sharing – he’s doing some of the best work I’ve seen in this area!
How is your “prompting cookbook” coming along? Any new and interesting learnings to share?
This week at Google IO, one of the projects covered was a new experimental one called Project Tailwind – see how Steven Johnson covered it on his Substack after the event. He’s been working part-time with Google on this project which he describes Tailwind this way:
Tailwind allows you to define a set of documents as trusted sources which the AI then uses as a kind of ground truth, shaping all of the model’s interactions with you. In the use case shown on the I/O stage, the sources are class notes, but it could be other types of sources as well, such as your research materials for a book or blog post. The idea here is to craft a role for the LLM that is not an all-knowing oracle or your new virtual buddy, but something closer to an efficient research assistant, helping you explore the information that matters most to you.
Google’s one line description is: “Tailwind is your AI-first notebook, grounded in the information you choose and trust.”
While working with the existing chatbots (ChatGPT, Google Bard, Microsoft Bing, etc.) is fun and useful, I’d be much happier having a research assistant which would primarily function on content I’ve created with an option to go beyond my content to the wider world. Johnson says he has “found that Tailwind works extremely well as an extension of my memory.”
Google’s initial implementation of Tailwind is based upon files in your Google Drive. For privacy reasons particularly, I’d especially welcome such a feature being trained and used locally on my own computer rather than having to upload my content to Google Drive and a cloud trainer.
I’ve requested access to Project Tailwind and look forward to experimenting with it when it’s made available. Meanwhile, here’s a short video that discusses Tailwind:
You must be logged in to post a comment.