The hottest Experiments Substack posts right now

And their main takeaways
Category
Top Business Topics

Experiments that never see the light of day

Briefly Bio • 19 implied HN points • 31 Oct 24
🔬 Science Experiments
  1. Many experiments go unpublished because they're too small or inconclusive. Even if they don't seem important, they really help build bigger discoveries.
  2. It's important for scientists to share these lesser-known experiments. Sharing can help the whole field of science progress faster.
  3. Open science encourages collaboration. Scientists and companies should talk to each other about new ways to share research.

Three Dumb Studies for your consideration

Experimental History • 19828 implied HN points • 10 Jun 25
🔬 Science Experiments
  1. Short and low-cost experiments can still provide interesting insights. Even simple studies can teach us something new.
  2. People often have unexpected reactions to pain and discomfort, like some even enjoy putting their hands in ice water. This shows that experiences can be more subjective than we think.
  3. Our preferences for things like sugar and salt are complex. People hesitate to eat them in pure forms due to social norms or taste expectations, indicating our relationship with these substances is nuanced.

MBTI, If You Want Me Back, You Need to Change Too

Granted • 8705 implied HN points • 05 Mar 23
🔬 Science Experiments
  1. Consider embracing the latest data to improve the theory by moving away from antiquated ideas and focusing on more reliable traits.
  2. Demand real evidence for efficacy by conducting randomized, controlled experiments to validate the benefits claimed by MBTI.
  3. Understand that many scientists have left MBTI due to concerns about its validity and its historical associations with racism and sexism.

How to Get Better At Learning From Experience

Range Widely • 1336 implied HN points • 07 Mar 23
🚌 Education Experiments
  1. Success is not always a good teacher, especially in unpredictable and impactful scenarios like earthquakes
  2. Differentiate between 'kind' and 'wicked' learning environments, where experience can either improve judgment or give false confidence
  3. Reflecting on and analyzing experiences can lead to better learning outcomes than relying solely on intuition or repetitive actions
Get a weekly roundup of the best Substack posts, by hacker news affinity:

Moving On

Software Design: Tidy First? • 1281 implied HN points • 10 Jan 25
🕹 Technology Experiments
  1. It's important to recognize when to move on from a project that isn't working. You don't have to stick with something just because you've already invested time or resources into it.
  2. Balancing between believing in your ideas and treating them as experiments is key. If something isn't getting good feedback, it's okay to change direction and try something new.
  3. Using timeboxing can help you make better decisions about projects. Setting a time limit lets you step back and reconsider if it's worth continuing or if you should explore other options.

Intrigue

Remote View • 334 implied HN points • 28 Jan 24
🔬 Science Experiments
  1. The post discusses the concept of intrigue both as a verb and a noun.
  2. The post announces a livestream on January 28, 2024, at 22:00 CET, discussing thoughts from a book by Robert Temple.
  3. The post provides various references, including books, articles, and research papers on topics like science, plasmoids, and cold war technology.

Exploring AI

Software Design: Tidy First? • 2496 implied HN points • 25 Jan 24
🕹 Technology Experiments
  1. Every technological revolution involves something valuable becoming cheaper.
  2. In an exploration phase, conduct many experiments and embrace duplication.
  3. Embrace the changed cost landscape and actively explore new possibilities.

A quarter-year update

Thought-gifts from Jasmine 🌱☁️ • 353 implied HN points • 28 Apr 23
🎭️ Culture Experiments
  1. Jasmine started a new company called Trellis and is very excited about its potential
  2. Jasmine has been traveling to different places and plans to settle down soon after signing a lease
  3. Jasmine is actively looking for a life partner and is interested in learning new skills like massage, silversmithing, and writing

Should we be nice to ChatGPT?

benn.substack • 613 implied HN points • 16 Jun 23
🕹 Technology Experiments
  1. ChatGPT performs better with neutral prompts than nice or mean tones.
  2. Being nice to ChatGPT can lead to more verbose responses and lower accuracy in completing tasks.
  3. Treating ChatGPT well or poorly is like a wager on its future impact, so choose wisely.

Erika Update #1: Getting projects started is hard

Erika’s Newsletter • 98 implied HN points • 20 Feb 23
🔬 Science Experiments
  1. Starting projects can be challenging, but perseverance and a supportive environment are key to making progress.
  2. Mistakes are common in research, but being able to identify and correct them is crucial for success.
  3. Regular updates and reflections on the progress of a project can provide valuable insights and contribute to overall success.

MK-ULTRA & The CIA's Bizarre Psychedelic Escort Service

Tripsitter • 99 implied HN points • 08 Jun 23
💾 History Experiments
  1. The CIA's MK-Ultra program involved using psychedelics for mind control and manipulation, leading to disturbing outcomes like forced confessions and personality alterations.
  2. Under MK-Ultra, experiments like Operation Midnight Climax and 'psychedelic driving' subjected individuals to involuntary drug doses, induced comas, and repeated messaging, resulting in severe negative effects on mental health.
  3. Conspiracies surrounding MK-Ultra include connections to infamous figures like the Unabomber, Charles Manson, and Jack Ruby, raising disturbing questions about the program's impact on society.

Open Friday Thread: Storytelling, Teaching Experiments, and Private Threads

The Recovering Academic • 98 implied HN points • 03 Feb 23
🚌 Education Experiments
  1. Collaboration on projects like Inner Life can help create a sense of community and engagement among contributors and subscribers.
  2. Storytelling and surprising teaching methods can have a profound impact on engagement and learning, both in traditional and experimental settings.
  3. Offering private threads or discussions for paying subscribers can enhance exclusivity and provide a more intimate space for in-depth conversations.

Embracing Experiments of Living | #284

Boundless by Paul Millerd • 121 implied HN points • 21 Dec 24
🎭️ Culture Experiments
  1. Trying new things in life can lead to personal growth and discovery. It's important to run experiments to learn what works for you.
  2. Reinvention is a gradual process, not a one-time event. As you change your behaviors, you may feel lost, but that's part of finding your new path.
  3. Being self-employed allows for flexibility and exploration. It's a way to actively engage with life, rather than just following the usual routines.

Being an Early Career Scientist

The Polymerist • 116 implied HN points • 06 Nov 24
🔬 Science Experiments
  1. Try new things in your career and treat them like experiments. If something doesn't work out, you can always adjust and try again.
  2. It's okay to feel uncertain at the beginning of your career. Each experience helps you learn more about yourself and what you want.
  3. Life will always bring challenges, but keep moving forward and experimenting. Enjoy the process of learning and discovering new possibilities.

Learning by Doing

Sunday Letters • 59 implied HN points • 15 Oct 23
🚌 Education Experiments
  1. Doing is often better than just thinking. If you want to learn a skill, just start practicing it a lot.
  2. Trying many things and being okay with failure helps you learn faster. When you're not focused on being perfect, you can improve more easily.
  3. Having a 'beginner's mind' can lead to better results. Beginners often do well because they aren't held back by too much prior knowledge or self-doubt.

More info, more problems

inexactscience • 39 implied HN points • 16 Nov 23
🔬 Science Experiments
  1. When people get more information, they often underreact instead of overreact. This means they might ignore new data instead of properly adjusting their predictions.
  2. Experiments showed that when faced with two variables, people made less accurate forecasts. Adding complexity actually made their predictions worse.
  3. Having clear instructions and understanding of the information really helps improve decision-making. If people are confused, they tend to ignore important details.

Seven Things the Duo Founders Did Right

Build To Scale • 19 implied HN points • 06 Feb 24
💼 Business Experiments
  1. High quality investors can significantly impact the growth and success of a company.
  2. Offering equity to all employees, not just top management, can drive a growth mindset and benefit everyone involved.
  3. Maintaining focus on core business, while being open to experimentation, can lead to sustainable growth and innovation.

16 Jan 2024 – Libyan Sands, germ warfare on the subway, and Boeing

Outsider Art • 19 implied HN points • 16 Jan 24
💾 History Experiments
  1. The book 'Libyan Sands' by Ralph Bagnold paints a poetic picture of the Libyan Desert and mentions the skeletons of camels marking certain routes.
  2. A study conducted by the US Army in 1968 on the vulnerability of subway passengers in New York City to biological agents showed how pathogens could be disseminated widely without detection.
  3. An unsettling quote from 'White Noise' by Don DeLillo offers a glimpse into the fear and chaos within an aircraft's cockpit during a crisis.

The Worst Outcome is a Mediocre Success

Pivotal • 154 HN points • 21 Sep 23
💼 Business Experiments
  1. Startup experimentation is about learning from clear successes or failures, not mediocre results.
  2. The danger of a mediocre success is the lack of clarity on whether a strategy can be scaled or needs to be changed.
  3. It's important to design experiments with well-defined outcomes to avoid the messy middle of mediocre success.

Time response of first order systems

Only Wonder Knows • 19 implied HN points • 27 Oct 23
🔬 Science Experiments
  1. When measuring a system's response time, make sure to wait for at least five time constants to reach a stable state.
  2. First order systems have a unique time response that can be calculated using simple expressions.
  3. Extracting the time constant of a system involves analyzing data, normalizing measurements, and ensuring a good fit to confirm the system's behavior.

Do Large Language Models have a "Reasoning Gap"?

The Irregular Voice • 2 HN points • 01 Apr 24
🕹 Technology Experiments
  1. Large Language Models (LLMs) may not always exhibit true reasoning abilities, with a potential reliance on memorization instead of learning general techniques.
  2. Synthetic data generation systems like MATH() can be used to explore the reasoning capabilities of LLMs, but may introduce biases if not carefully analyzed and corrected for errors.
  3. Fine-tuning LLMs on specific problem areas can reveal insights into their reasoning abilities, but challenges with longer solutions and complex problem sets may impact performance.

Where do LLMs spend their FLOPS?

Artificial Fintelligence • 13 implied HN points • 29 Jan 24
🕹 Technology Experiments
  1. FLOPS in LLMs are mainly spent on computing QKV, attention output matrix, and running the FFN.
  2. Wider LLM models parallelize better and favor lower latency, while deeper models linearly increase inference time.
  3. Empirical analysis shows linear scaling in performance as LLM model dimensions increase.