Hello, AI Fans!

This week we address Google’s launch of its new experimental reasoning AI model, Gemini 2.0 Flash Thinking, aimed at multimodal understanding and complex problem-solving and look into AI pioneer Geoffrey Hinton’s warning of a 10-20% chance of human extinction within 30 years due to rapid and potentially uncontrollable artificial intelligence development!

This week’s edition:

• Google Launches Its Own Reasoning AI Model: Gemini 2.0 Flash Thinking

• AI Pioneer Raises Likelihood of Technology Causing Human Extinction Within 30 Years

• Your 5 min AI learning - Get Podcast Summaries Straight to your Inbox with Podsift

• New AI tools

• More AI News Highlights

• AI Image of the Week

Read time: 8 minutes

LATEST NEWS

GEMINI 2.0 FLASH THINKING

Google Launches Its Own Reasoning AI Model: Gemini 2.0 Flash Thinking

What You Need to Know: Google has released a new experimental AI model called Gemini 2.0 Flash Thinking Experimental, focusing on reasoning capabilities. The model is available on Google's AI Studio platform and is designed for multimodal understanding, reasoning, and coding. The model is built on the Gemini 2.0 Flash model, similar to OpenAI's o1 and other reasoning models.

The Finer Details:

• The model pauses before responding, considering related prompts and explaining its reasoning.

• It aims to fact-check itself, potentially avoiding common AI pitfalls.

• Reasoning models typically take longer to arrive at solutions, from seconds to minutes.

• Early testing shows room for improvement, with some inaccuracies noted.

So What? This development reflects a broader industry trend towards more sophisticated AI reasoning capabilities, with Google reportedly dedicating substantial resources, including multiple teams and at least 200 researchers, to this technology. The industry sees reasoning models as a potential solution to the diminishing returns of traditional AI scaling techniques, offering new approaches to refine generative AI. However, challenges remain, including high computational costs and uncertainty about long-term progress potential.

HUMAN EXTINCTION CAUSED BY AI?

AI Pioneer Raises Likelihood of Technology Causing Human Extinction Within 30 Years

What you Need to Know: Geoffrey Hinton, often called the "godfather of AI," has increased his estimate of the likelihood that AI could lead to human extinction within the next 30 years.

The Finer Details:

• Hinton now estimates a 10% to 20% chance of AI causing human extinction, up from his previous 10% estimate.

• He attributes this increased risk to the unprecedented challenge of managing entities more intelligent than humans.

• Hinton, who was awarded the Nobel Prize in Physics this year for his work in AI, expressed these concerns in an interview with BBC Radio 4.

• He emphasised that humanity has never before dealt with things more intelligent than ourselves.

So What? Geoffrey Hinton's increased estimate of AI extinction risk highlights the urgent need for robust safety measures and global cooperation in AI development. His warning suggests AI capabilities may be advancing more rapidly than expected, emphasising the challenge of balancing innovation with safeguards against potential catastrophic outcomes. This heightened concern is likely to spark increased public discourse and education about AI risks and benefits, underscoring the importance of addressing AI safety as a global priority.

YOUR 5 MIN AI LEARNING

Get Summaries of your Favourite Podcasts with Podsift

For this week's AI learning section, we're excited to introduce you to Podsift, the perfect tool for catching up on all those podcasts you've missed! If like us, you have a long list of podcasts that you’ve been meaning to listen to but never seem to find the time for (which happens to the best of us!), this tool might be the perfect fit for you. Podsift is a free service that sends AI-generated summaries of your favourite podcasts to your inbox. Signing up is a breeze: just enter and verify your email, choose the podcasts you'd like summaries for, and voilà! You'll receive concise episode summaries every time a new episode drops. It's a fantastic way to stay informed without the time commitment of listening to every single episode—perfect for those of us who have busy schedules! 

So the next time you’re left wondering what you missed in this week’s episode, why not try Podsift and make your podcast experience more efficient?

LATEST AI TOOLS

TimeSentry

A tool for time-tracking and billing for professionals (Freemium)

“TimeSentry provides all the basic features and benefits of existing time tracking and billing software while adding a meta-layer of AI to auto populate timesheets; so your firm can spend less time billing and more time billing.”

Backflip AI

A tool to turn ideas into 3D-printable models (Paid)

“Turn your ideas into real parts — today. Create at the speed of your imagination, with AI 3D design tools for the physical world.”

SiteRails

A tool to automate social media content creation and management (Paid)

“Automated Social Presence For Your Business. Create months of content from your existing website, or craft posts manually.”

MORE AI HIGHLIGHTS

AI Tool Identifies Heart Condition Prior to Symptom Onset

A new AI tool developed by the University of Leeds can detect atrial fibrillation (AF) before symptoms appear, potentially reducing the risk of deadly strokes. By analysing GP records for "red flags," the tool identifies individuals at risk based on factors like age, sex, and existing medical conditions. Currently undergoing a six-month trial in West Yorkshire, those flagged as high-risk will use a handheld ECG device to monitor their heart rhythm. If successful, this could lead to a nationwide rollout, significantly improving early detection and treatment of AF, which contributes to around 20,000 strokes annually in the UK.

OpenAI Sponsors a $1 Million Research Initiative at Duke University Focused on AI and Morality

OpenAI has awarded a $1 million grant to Duke University to study how AI can predict human moral judgments. Led by ethics professor Walter Sinnott-Armstrong, the “Making Moral AI” project aims to develop a “moral GPS” to assist in ethical decision-making across fields like medicine and law. The research will explore how moral attitudes are formed and the challenges of integrating the emotional and cultural nuances of morality into AI. This initiative emphasises the importance of aligning AI tools with societal values, ensuring responsible innovation in technology.

Groundbreaking AI Camera Aims to Combat Drunk Driving

A groundbreaking AI camera is being tested in Devon and Cornwall to identify drivers under the influence of drugs or alcohol. The innovative Heads-Up machine can detect behaviours indicative of impairment, allowing police to stop vehicles for roadside testing. Developed by Acusensus, this camera can be quickly relocated to various roads without prior notice, enhancing its effectiveness in catching impaired drivers. With drunk drivers being significantly more likely to cause fatal accidents, local authorities hope this technology will improve road safety. The trial is taking place throughout December, coinciding with other anti-drink-driving campaigns, as part of a larger commitment to reducing fatalities on the roads.

PwC is Hosting "Prompting Parties" to Train Employees in Using AI Tools Within a Relaxed Environment

PwC is addressing the AI skills gap by hosting "prompting parties" - group sessions where employees can safely experiment with generative AI tools like Microsoft Copilot and ChatPwC. Since launching in March, they've hosted nearly 500 parties with over 880 more requested, aiming to help employees become comfortable using AI in a low-stakes, collaborative environment. These sessions allow staff to learn AI prompting techniques, explore real use cases, and reduce anxiety about using AI in professional settings.

AI IMAGE OF THE WEEK

As we approach the New Year, we couldn't resist creating a festive image to capture the spirit of celebration! We wanted to see if AI image generators could encapsulate the joy and excitement of ringing in the New Year, reminiscent of those enchanting moments we cherish with loved ones.

For our prompt, we crafted something special using Grok:

“Create a vibrant New Year’s celebration scene featuring a group of friends joyously counting down to midnight. The setting should be adorned with colourful fireworks lighting up the night sky, surrounded by sparkling decorations and a lively atmosphere filled with laughter and joy. The illustration should evoke a sense of warmth and togetherness, capturing the magic of new beginnings.”

Send us your AI generated art for a chance to be featured.

FEEDBACK

• Let us know how we did - shoot us a quick email. We read every single one (no AI in sight!).

• If you enjoyed this newsletter please share with an awesome friend or two.

• If you are that awesome friend then hit the below button: