SafetyTech News | Issue 035

Welcome back to Issue 35 of Safety Tech News - let's dive straight in.

A few words on AI x Safety

First things first. It's okay if you are totally confused about AI and it's role in health and safety - join the club. AI is in every #safetytech product on the market is a wild space and its moving super fast so here's a few things I'd like to note with the current state of play (mid-Feb 2024).

1) You need base level digital literacy to join the conversation: Search "IBM + AI" in your favourite browser and it should give you hits on IBM's intro to AI content - it's my go-to entry level resource. Even better, head over to Perplexity (it's free and awesome, if you've never used it - thank me in the comments) use this prompt "I'm completely new to AI can you please provide me with a a short bullet point list overview of key terms and a short list of entry level resources (free courses, journal articles, blogs and news) I can review" it will deliver you the goods. Read up. Soak.

2) Use case exploration: If you've completed 1) you'll know AI is an umbrella term for several different technologies, mostly machine learning. So with that in mind, in my view, there are currently 4 key use case buckets applying AI for health and safety:

1. Generative AI: using solutions like ChatGPT and M365 Copilot for health and safety administrative tasks. Results are mixed but getting better - massive need for human in the loop
2. Advanced Analytics: using AI to identify trends and insights for action from work and safety data - unfortunately results at this stage are hamstrung by the poor quality of data. We have MUCH work to do in the data space if we want to leverage the potential here.
3. Conversational AI Assistants: this is using natural language processing + gen AI to have a conversational interface with your management systems. Things are progressing nicely in this space but it's still early days - humans are SUPER important here too.
4. Computer Vision: machine learning over video footage - well established, great results but hinges on strategy and change management + trust, cyber, ethics etc.

3) Never stop learning: Things are moving VERY fast so if you tried something and formed an opinion on it last month - best to reserve judgement and try again.

I'll be diving into these few topics plus more in a webinar later this week:

You can sign up here...

That leads me to two super interesting advancements in the world of AI this week.

OpenAI released Sora: Sora is an AI model capable of generating realistic and imaginative videos from text instructions. It can create videos of up to 60 seconds, featuring detailed scenes, complex camera motion, and multiple characters with vibrant emotions. Sora is a diffusion model with a transformer architecture, similar to the GPT models, and it can understand and simulate the physical world in motion. The model has a deep understanding of language, enabling it to accurately interpret prompts and generate compelling characters that express vibrant emotions. Sora is currently being tested by security researchers and a small group of visual artists, designers, and filmmakers to assess its safety and gather feedback on its capabilities. OpenAI has not yet announced a public release date for Sora and is working on implementing important safety measures before integrating it into their product

Google released Gemini 1.5: Google Gemini 1.5 is an upgrade to Google's flagship AI model, which is a large language model (LLM) like OpenAI's GPT 4. It delivers dramatically enhanced performance, with a breakthrough in long-context understanding across modalities. Gemini 1.5 Pro, the first version being made available, comes with a standard 128,000 token context window, but Google is testing a context window of up to one million tokens. Note: Tokens are the smallest building blocks used by AI models to process information. In the context of AI language models, a token can be an entire word, part of a word, or a subsection of a word. Gemini 1.5 can process vast amounts of information in one go, including 1 hour of video, 11 hours of audio, codebases with over 30,000 lines of code, or over 700,000 words. It also introduces a breakthrough experimental feature in long-context understanding, allowing it to learn a new skill from information given in a long prompt without needing additional fine-tuning.

What do these mean for safety? For now I don't really know, but it's these developments have certainly piqued my curiosity... the foundational mindset needed to leverage tools like Sora & Gemini.

Our friends at Everyday Massive. The human experience co. have a few things to say on the topic... let's explore.

Curiosity with Everyday Massive

If you haven't heard of Everyday Massive - stop scrolling and go check out their website, connect with their founder Jen Jackson - she's also the author of a fabulous book How to Speak Human (recommended muchly). Their musings on curiosity are worth a read.

Now with a curious mind - let's further explore the world of AI x Safety.

NVIDIA | Using the power of AI to make factories safer

英伟达 are responsible for designing and manufacturing graphics processing units (GPUs), as well as for its contributions to artificial intelligence (AI) hardware and software. They provide excellent technical blogs and they recently provided an overview of how their solutions can be leveraged to enhance safety in factories.

Check out the full blog below.

Speaking of NVIDIA - I'm super jealous of Prakash Senghani CEO of Navatech the team behind Saifety.ai who got to meet Jensen! Total #safetytech nerd moment... Check out the post:

EQ for AI: shaping tomorrow’s human-machine experience

Siemens publish a newsletter on LinkedIn - its worth a follow for nuggets of wisdom like this.

"the future of human-machine collaboration is not just about form and function. It is about experience too – the extent to which users can positively engage with their digital companions, communicate their intentions, and feel that a genuine connection has been established. This connection is the catalyst for successful collaboration; a shared understanding to support co-creation"

Read the article here

moving on...

intenseye Hot Object Detection Model

intenseye 's CEO Sercan Esen recently posted a new use case to their portfolio - Hot Object Detection. Check it out here:

Construction Innovation with AtkinsRéalis

Innovation in the construction sector is hard work... check out how AtkinsRéalis making progress with Lava Labs.

+ more here

Valeria Spirovski on "Making Space for Design"

I was delighted to see this post pop up on my feed - it's foundational for those seeking to improve the design of work, particularly if you're seeking to leverage technology in the process. Click the link for the goodness - make sure you follow Valeria Spirovski for more.

Making space for Design - Defining Design, why it’s so important and introducing a unifying model of design

Protex AI On Demand Webinar: Your 2024 Safety AI Project | Journey From Implementation To Success

Larry J. Newport shares his personal experience planning the implementation of Protex AI .

NAEM | Tech24 Conference

The NAEM Tech24 | EHS + ESG Data & Reporting Conference is now open for registration - more information here

Ava - Avettas AI-Assistant is here

Avetta has announced their AI assistant - details are limited and I haven't had a chance to test it our so I have no further info for you unfortunately. Best to check out their press release, teaser video and further information here:

AI + HI Project

The AI+HI Project is designed to provide an intimate and interactive experience focused on innovation and forward-thinking collaborations between Artificial Intelligence and Human Ingenuity in the workplace.

HFESA Conference announced & the theme is excellent

The Australian HFESA Human Factors and Ergonomics Society of Australia have announced the theme and location of the 2024 conference and it's a real beauty. You can learn more about the conference here and contact Sara Pazell, PhD, CPE for sponsorship opportunities - there's some great options at reasonable costs.

Ok that's a wrap for this issue but before you go... a gentle reminder to DM or tag Cam Stevens for posts or news you'd like featured in Safety Tech News. Also - remember, feedback is a gift - please like, comment, share, follow and engage so we know what you want more or less of.

Stand by for Issue 036 but for now.

Be Safe, and Stay Curious.

Cam Stevens