Google Gemini Surpasses ChatGPT to Claim The #1 Spot 🏆

Welcome to another edition of Horizon AI,

The competition between OpenAI and Google is intensifying, and this time, it looks like Google has taken the lead with its new model.

Let’s jump into it!

Read Time: 4.5 min

Here's what's new today in the Horizon AI

• Google's new Gemini model Surges to the No. 1 Spot in the Chatbot Arena 🥇

• AI-Enhanced Mammography Boosts Cancer Detection by 21%

• AI Tutorial: How to prevent AI from using your data for training

• AI Tools to check out

• The Latest in AI and Tech 💡

• AI Findings/Resources

AI News

GOOGLE

Google's new Gemini model Surges to the No. 1 Spot in the Chatbot Arena 🥇

Google’s latest experimental model, Gemini-Exp-1206, has claimed the top spot in the Chatbot Arena leaderboard, overtaking OpenAI’s ChatGPT-4o.

Details:

• Gemini-Exp-1206 is the No. 1 model overall. It ties with OpenAI’s model in the coding category but leads in nine other categories, including instruction following and math.

• This is not the first time Gemini has taken the #1 spot; the Gemini-Exp-1114 model had already managed to do so until it was dethroned by ChatGPT-4.0 (20241120) on Nov. 21st. The models from both these companies seem to be taking turns at topping the AI benchmarks.

• The model has an impressive 2 million token context window and can analyze video content, a feature OpenAI and other competitors have yet to implement.

The model is accessible for free to everyone through Google AI Studio and the Gemini API.

Source

TOGETHER WITH GAMMA

The future of presentations, powered by AI

Gamma is a modern alternative to slides, powered by AI. Create beautiful and engaging presentations in minutes. Try it free today.

AI RESEARCH

AI-Enhanced Mammography Boosts Cancer Detection by 21%

A U.S. breast-screening program has highlighted the potential benefits of incorporating AI into mammography. The findings were shared at the Radiological Society of North America (RSNA) annual meeting.

Details:

• A study involving 747,604 women demonstrated that AI-assisted mammograms acted as a “second set of eyes” for radiologists, helping to identify anomalies.

• Women who used AI-enhanced scans had a 43% higher overall cancer detection rate, although 22% of this was attributed to selection bias (higher-risk patients were more likely to enroll in this self-pay program).

• The remaining 21% boost in cancer detection is credited to AI’s ability to improve recall rates for further imaging.

While the findings are promising, the researchers emphasize the need for further studies, such as randomized controlled trials, to better quantify AI’s impact on improving breast cancer detection, since More than a third of women across 10 health care practices chose to enroll in a self-pay, artificial intelligence (AI)-enhanced breast cancer screening program,

Source

AI Tutorial

How to prevent AI from using your data for training

AI tools often use your inputs to train their models. While some AI companies anonymize this data, you might be uncomfortable with this. Fortunately, most offer settings to disable this training.

Here’s how to turn off AI training on some of the most popular apps:

1. ChatGPT

On the web:
Click your profile picture (top right), go to Settings > Data control, and switch off the 'Improve the model for everyone' toggle.

Using the mobile app:
Open the menu (top left), tap the three dots next to your account name, then go to 'Data controls' to find the 'Improve the model for everyone' toggle.

2. Gemini

For Gemini, you need to turn off chat history to stop your conversations from being used for training.

On the web:
Click 'Activity' (bottom left).

Using the mobile app:
Tap your profile picture (top right), then 'Gemini Apps Activity.'

In both cases, you’ll find a toggle or button on the next screen to turn this feature on or off.

Even with activity logging turned off, chat data is kept for 72 hours. Starting a new chat will notify you that the current session will be deleted.

3. Copilot

On the web:
Click your profile picture (top right), then select your name and go to 'Privacy.' You'll find two toggles: 'Model training on text' and 'Model training on voice'—both can be turned off.

Using the mobile app:
The same toggles are in the mobile app. Tap your profile picture (top right), then Account > Privacy.

4. Grok

On the web:
Go to More > Settings and privacy > Privacy and safety > Grok & Third-party Collaborators, and uncheck the box.

Using the mobile app:
Tap your profile picture (top left) to access the same 'Settings and privacy' menu.

5. Perplexity

On the web:
Click the gear icon near your username (bottom right), then turn off the 'AI Data Retention' toggle under 'Account' to stop your data from being used in training.

Using the mobile app:
Tap your profile picture (top left), then locate the 'AI Data Usage' toggle.

6. LinkedIn

On the web:
Click your profile picture (top of the page), go to Settings & Privacy > Data privacy, and switch off the 'Data for generative AI improvement' toggle.

Using the mobile app:
Tap your profile picture (top left), then go to Settings > Data privacy > Data for generative AI improvement to find the toggle.

For other apps you might be using, the process should be similar. However, some apps like Meta AI and Reddit don’t offer an opt-out option within the app, so by using them, you’re implicitly accepting that your data may be used to train models.

AI Tools to check out

✌️ Twos: Organize notes, to-dos, reminders, events, and more with this AI-powered tool.

🗣 Wispr Flow: Speak naturally, write perfectly & 3x faster in every app.

🎮 Charmed: Offers AI-powered video game design tools.

🏃 Olyup: Monitor, evaluate, and enhance your fitness goals through a sports performance AI chatbot.

📊 Tables by Playmaker: Transform unstructured data into actionable data, instantly

Whenever you are ready, here's how we can help you:

Book an Audit call with our AI Automation agency: Our clients saved 40+ hours in their first month. Limited spots available. (Your business must be making a minimum of $5K/month)

AI Findings/Resources

🔍 OpenAI’s new ad shows ‘Reasoning’ AI making basic errors

🦾 Enterprise Generative AI workflows you can use today

💡 5 ways AI can help crisis response around the world

🤔 What Anthropic's AI knows about you

The latest in AI and Tech

X pulls new Aurora image generator

X briefly added the Aurora image generator to Grok, but it vanished a day later, replaced by FLUX beta. Aurora impressed with its photorealistic images and being largely unrestricted.

This came just a few days after X announced that users would no longer need to pay for X Premium to use Grok.

New Google AI weather model beats most reliable forecast system

GenCast, Google's new AI model, just outperformed the world’s leading weather forecasting system, according to a research paper published this week.

Reddit starts testing an AI-powered answers feature

Reddit is trying out a new AI feature to help users quickly find answers based on user posts. Only a small group in the U.S. can test it for now.

Meta launches Llama 3.3 AI model

Meta’s new Llama 3.3 70B model matches the performance of its larger predecessor but at a lower cost.

That’s a wrap!

Thanks for sticking with us to the end! Let’s stay connected on LinkedIn and Twitter.

Not subscribed yet? Sign up here and send it to a colleague or friend!

See you in our next edition!

Gina 👩🏻‍💻