😯 Grok did an oopsie

🧵 In today's edition:

• 🤖 OpenAI previews their newest model

• 😯 Grok did an oopsie

•  🧬 Chai time

• 🛠️ AI Tools to Check Out

• 🤑 AI Fundraising News

• 🐦 Tweet Post of the Day

🤖 OpenAI previews their newest model

Huge news coming out of OpenAI HQ.

OpenAI unveiled a series of AI models called OpenAI o1, designed to excel at complex reasoning tasks. These models are trained to think more thoroughly before responding, similar to human problem-solving processes.

Here’s the TL;DR:

• Impressive performance in science, coding, and math.

• Significantly outperforms previous models in challenging tasks like the International Mathematics Olympiad.

• Enhanced safety features, with improved resistance to "jailbreaking" attempts.

• Collaboration with U.S. and U.K. AI Safety Institutes for evaluation and testing

OpenAI is also introducing o1-mini, a faster and more cost-effective version optimized for coding tasks. It's 80% cheaper than o1-preview, making it ideal for applications requiring reasoning without extensive world knowledge.

Here is the current availability schedule:

• ChatGPT Plus and Team users: Immediate access with weekly rate limits.

• ChatGPT Enterprise and Edu users: Access beginning next week.

• API access: Available for qualifying developers (Tier 5).

• There are also plans to extend o1-mini access to all ChatGPT Free users.

Overall: While o1-preview currently lacks features like web browsing and file uploads, OpenAI plans to add these capabilities in future updates. The company will continue developing the GPT and o1 model series to cater to diverse user needs.

Read more: OpenAI Blog

😯 Grok did an oopsie

X's AI chatbot, Grok, recently spread misinformation about ballot deadlines after Joe Biden withdrew from the 2024 presidential race.

Here’s what happened: 

• Grok mistakenly claimed that new candidates couldn’t be added to ballots in nine states, sparking concern among election officials. 

• A group of state secretaries quickly flagged the error to X, prompting the company to revise Grok's responses. 

• After some initial pushback, X agreed to direct users to official voting websites when asked about election-related topics.

But this incident isn’t all that’s problematic about Musk’s chatbot: 

• Grok, marketed as an "anti-woke" chatbot, has raised broader concerns about its vulnerability to misinformation, likely due to its design and data sources. 

• There are also worries about its image-generation capabilities, which could potentially create misleading or inflammatory content.

The takeaway? This incident revealed two key points:

1. Negative: It exposed the risks of AI spreading election misinformation and the challenges for election officials.

2. Positive: It highlighted the importance of swift intervention, as the successful actions of election officials showed how public pressure can correct AI missteps.

With the 2024 election nearing, vigilance and cooperation between tech companies and election authorities will be essential in ensuring accurate electoral information.

Read more: Guardian

 🧬 Chai time

If you’ve been a long-time reader of NAB, you know that we’ve always made one big prediction about which industry will see the most transformative impact from AI. 

Spoiler: it’s healthcare.

And it feels like every few weeks, we’re proven right.

This week, Chai Discovery unveiled Chai-1, a groundbreaking multi-modal foundation model for predicting molecular structures. 

Chai-1 excels in key drug discovery tasks, providing unified predictions for proteins, small molecules, DNA, RNA, and covalent modifications.

Here’s the TL;DR:

• Chai-1 delivers state-of-the-art performance across benchmarks like PoseBusters and CASP15 protein structure prediction.

• Unlike many existing tools, it operates without multiple sequence alignments while maintaining high accuracy.

• It outperforms previous models in predicting multimer structures, even using just single sequences.

• Chai-1 can integrate new data, like lab-derived restraints, to enhance performance.

• The model is freely accessible via a web interface (including for commercial use), with weights and code available for non-commercial use.

Overall: Chai Discovery, founded by experts from top AI and drug discovery firms, aims to revolutionize biology by making it an engineering discipline. 

The team plans to continue developing AI models to predict and reprogram biochemical interactions.

Read more: Chai Discovery

🛠️ AI Tools to Check Out

• Copper: Copper helps you connect with leads, win deals, deliver projects, and create repeat clients. All in one easy-to-use tool. Check it out!

• Moonhub: Moonhub's world-class talent experts deploy proprietary AI to help you hire top candidates faster. Check it out!

• SoBrief: Turn any book into an audiobook in 10 mins. Supported in 40 languages. Check it out!

🤑 AI Fundraising News

• Cavela raises $2M in funding to provide a platform that enables brands and businesses to source products. It uses AI-driven automation to connect them with premium manufacturers worldwide at optimal cost.

• Jus Mundi raises $22M in Series B funding to provide global legal intelligence, data-driven arbitration professional selection and business development services.

🐦 Tweet Post of the Day

Devin teamed up with OpenAI on the launch of their latest model - more details below 💪