The hottest Safety Substack posts right now

And their main takeaways
Category
Top Technology Topics

Prep for the Vaccine

Are You Okay? • 0 implied HN points • 20 Jan 21
🏥 Health & Wellness Safety
  1. More COVID-19 vaccine doses are on the way, including a safe single-dose option from Johnson & Johnson.
  2. Stay patient with booking vaccine appointments, as more doses are being distributed, and federal efforts are underway to speed up the process.
  3. Prepare for your vaccine appointment by ensuring you're well-rested and well-fed, but avoid premedicating with Benadryl if you have allergies.

Hit the Road

Are You Okay? • 0 implied HN points • 16 Jun 20
🏥 Health & Wellness Safety
  1. Know where you are going and consider the COVID-19 situation in that area to plan your road trip accordingly.
  2. Prepare by mapping out your travel route, filling up your gas tank before departure, and packing snacks for the journey.
  3. Be cautious during overnight stays in hotels, bring cleaning supplies, wash glassware before using, and opt for stairs over elevators when possible.

Part 4: Ways to reduce noise when assessing school shooting threats

School Shooting Data Analysis and Reports • 0 implied HN points • 16 May 24
🏥 Health & Wellness Safety
  1. Having groups assess school shooting threats leads to more consistent and accurate judgments by reducing individual variability.
  2. Utilizing the wisdom of crowds, where multiple assessments are averaged, can help in decision-making and improve accuracy, especially in critical situations like school shooting threats.
  3. Implementing algorithms for threat assessment, alongside human judgement, can standardize evaluations, reduce bias, and potentially enhance decision-making processes in school safety.

Film Study: Learning from North Kansas City High basketball game shooting

School Shooting Data Analysis and Reports • 0 implied HN points • 09 Apr 24
🚌 Education Safety
  1. After a shooting incident, school officials should analyze security plans and procedures by studying the events similar to how coaches review game film.
  2. Many schools lack specific plans or training for shootings during sports events, highlighting the need for better emergency coordination, notification systems, and radio interoperability.
  3. Security technology like acoustic gunshot detection systems and CCTV image classification software can provide valuable insights, but their effectiveness relies on proper implementation and resolution of cameras.

The ‘Matrix Problem’ for School Security

School Shooting Data Analysis and Reports • 0 implied HN points • 23 Mar 23
🚌 Education Safety
  1. The 'Matrix Problem' in school security occurs when staff are unprepared to handle an armed assailant at security checkpoints or during searches.
  2. Success in addressing school shooting threats hinges on staff having formal training, appropriate equipment, and taking the threat seriously.
  3. Ensuring school security requires proper training and equipment for staff, avoiding forcing unprepared personnel into security roles.
Get a weekly roundup of the best Substack posts, by hacker news affinity:

AI catastrophes and rogue deployments

Redwood Research blog • 0 implied HN points • 03 Jun 24
🕹 Technology Safety
  1. Categorizing AI catastrophes based on rogue deployments or lack thereof is crucial for understanding safety measures and potential threats.
  2. Rogue deployments, where safety measures are absent, pose a significant risk as they make causing a catastrophe much easier compared to non-rogue deployments.
  3. Different attacker profiles like scheming AIs, lab insiders, and external attackers play a key role in determining the type and severity of AI-related catastrophes.

Untrusted smart models and trusted dumb models

Redwood Research blog • 0 implied HN points • 07 May 24
🕹 Technology Safety
  1. The most reasonable strategy to assess if AI models are deceptively aligned is to test their capability; incompetent models are less likely to be deceptively aligned.
  2. By using capability evaluations, models tend to fall into categories of untrusted smart models and trusted dumb models.
  3. Combining dumb trusted models with limited human oversight can help mitigate the risks posed by untrusted smart models.

Can the Anti-Parasite Medication Ivermectin Treat COVID-19? Probably Not

Harnessing the Power of Nutrients • 0 implied HN points • 17 Apr 20
🏥 Health & Wellness Safety
  1. Ivermectin may not be effective in treating COVID-19 based on research findings in isolated cells versus actual blood concentrations needed for effectiveness.
  2. Sustaining the necessary antiviral concentration of ivermectin in the blood for 48 hours is not feasible with current dosing regimens and could lead to unsafe levels of the drug.
  3. It is crucial to rely on verified medical advice and research when considering treatment options for COVID-19, as unsubstantiated approaches like ivermectin may pose risks without proven benefits.

Ngồi ở đâu trên máy bay là an toàn nhất?

Thái | Hacker | Kỹ sư tin tặc • 0 implied HN points • 21 Jul 07
🏝 Travel Safety
  1. Ngồi càng gần phía đuôi máy bay sẽ giúp tăng cơ hội sống sót trong trường hợp tai nạn máy bay
  2. Vị trí ghế trên máy bay cũng đóng vai trò quan trọng trong việc đảm bảo an toàn cho hành khách
  3. Thông tin về vị trí an toàn trên máy bay có thể giúp hành khách cảm thấy yên tâm khi đi du lịch

GPT-V Coming Soon?

Sector 6 | The Newsletter of AIM • 0 implied HN points • 21 Dec 23
🕹 Technology Safety
  1. OpenAI is likely to release GPT-5, also known as GPT-V, sooner than expected due to safety concerns with advanced AI systems.
  2. They have launched a new grant program to support research focused on making AI safer and more aligned with human values.
  3. There's a new preparedness framework in place to help anticipate and protect against potential risks of more powerful AI models.

The Fault in Our Stars

Sector 6 | The Newsletter of AIM • 0 implied HN points • 06 Apr 23
🕹 Technology Safety
  1. OpenAI has faced criticism and calls for a pause on its AI experiments, particularly around powerful models like GPT-4.
  2. The company claims to be working on developing safe and responsible AI systems, but many experts doubt the effectiveness of this approach.
  3. Some critics view OpenAI's recent communication as just a public relations effort without real substance.

AI Engineer World's Fair–Day 2: From Beats to Bytes

Code and Context • 0 implied HN points • 29 Jun 24
🕹 Technology Safety
  1. AI is making big waves in both creative fields like music and practical industries. It's changing how people create and work every day.
  2. Companies are figuring out how to trust AI and keep things safe, showing the importance of balancing innovation with responsibility.
  3. Lessons learned from platforms like Discord show how to successfully integrate AI into existing structures and improve experiences for users.

How Should Large Language Models Be Evaluated?

Cobus Greyling on LLMs, NLU, NLP, chatbots & voicebots • 0 implied HN points • 01 Nov 23
🕹 Technology Safety
  1. Large Language Models (LLMs) should be evaluated based on their knowledge, alignment, and safety. This helps ensure they meet necessary standards.
  2. Evaluation has become more complex as LLMs can do higher-level tasks, rather than just basic language checks like syntax and vocabulary.
  3. Creating a clear taxonomy for LLM evaluation helps guide researchers and companies in assessing these models effectively.

Aerially Separated Nuclear Aircraft

Tapa’s Substack • 0 implied HN points • 06 Jul 24
🕹 Technology Safety
  1. Nuclear aircraft need heavy shielding, which can make them less efficient. The idea is to separate the nuclear reactor from the passenger plane using a drone.
  2. The drone would beam power to the airplane using lasers, which could reduce the need for heavy shielding and keep passengers safer.
  3. This concept suggests that enough laser power can be delivered to make flying possible, but it's still just an idea and might not work out.

The 2024 Creta is Perfection

Kartick’s Blog • 0 implied HN points • 10 Dec 24
🕹 Technology Safety
  1. The 2024 Creta is very comfortable and spacious. It offers a high seating position, plenty of legroom, and smooth suspension, making drives feel enjoyable without bumps.
  2. Driving the Creta is effortless and fun. Its powerful turbo engine and light steering make it easy to handle, even in heavy traffic, and it feels stable at high speeds.
  3. The car has great digital features like wired CarPlay and a premium sound system. However, some tech issues exist, like non-multitouch screens and limited app functionalities.

Toyota Fortuner: Buy Scorpio N Instead

Kartick’s Blog • 0 implied HN points • 01 Jan 25
🕹 Technology Safety
  1. The Toyota Fortuner is a true SUV, but it costs over double the price of the Scorpio N, which offers similar features and capabilities.
  2. Both vehicles have the same seating and safety features, but the Fortuner doesn't provide enough comfort or modern design to justify its higher cost.
  3. Choosing the Scorpio N could save you money, allowing you to afford additional vehicles or upgrades while still getting a capable SUV.

The Kia Carens Is the Best 3-row Vehicle

Kartick’s Blog • 0 implied HN points • 25 Dec 24
🕹 Technology Safety
  1. The Kia Carens is considered the best 3-row vehicle, especially for those needing extra space.
  2. It offers useful features like ventilated seats and a cooled cup-holder, but has some comfort issues like limited legroom in certain seats.
  3. If you don't require a 3-row vehicle, options like the Creta may provide more comfort and features, making them a better choice.

The Toyota Hyryder / Grand Vitara is actually a Mediocre Vitara

Kartick’s Blog • 0 implied HN points • 20 Dec 24
🕹 Technology Safety
  1. The Toyota Hyryder and Grand Vitara have poor comfort and driving features compared to competitors like the Creta and Seltos. Many aspects like seat comfort and engine sound were disappointing.
  2. The car has some good features like various cameras and a remote AC start, but it lacks important safety features like side airbags and lane departure warning. This makes it feel incomplete for the price.
  3. Overall, both the Hyryder and Grand Vitara feel similar but offer limited benefits. Buyers may find better options in other cars for the same price.

55. Puerto Vallarta Uncovered: A Coastal Paradise in Mexico

Spyderdoc Newsletter • 0 implied HN points • 12 Jul 25
🏝 Travel Safety
  1. Puerto Vallarta is a beautiful coastal town in Mexico with great weather, nice beaches, and delicious food. It's a popular destination for travelers looking for relaxation and fun.
  2. The culture in Puerto Vallarta is vibrant and welcoming, with friendly locals and a mix of expats from Canada and the U.S. who have made it their home.
  3. Activities in the area range from exploring local markets and enjoying artsy boardwalks to swimming in the ocean and taking in stunning sunset views. There's something for everyone here!

🐿️ Don’t Worry, Nobody’s On Fire - The Insanely Profitable Tech Newsletter

Squirrel Squadron Substack • 0 implied HN points • 09 Feb 26
🕹 Technology Safety
  1. Using passive language in reports hides who actually did what and makes it hard to hold anyone accountable.
  2. Paperwork and process fixes are useful but not enough; if root cause analysis ignores human mindsets and norms, the same failures will recur.
  3. Leaders need to watch how people really behave, name specific actions and responsibilities, and enforce accountability to change harmful cultural habits.