Google’s Gemini Is Finally Out, But How Does It Compare To ChatGPT? 💠 

• Gemini: Google’s largest and most capable AI model

• Elon Musk Releases Generative AI Chatbot Grok on X

• EU Agrees ‘Historic’ Deal with World’s First Laws to Regulate AI

• Weekly Productivity Hacks

• Roast Profiles ☕️ 

• Quick Snips of Other AI News

Read time: 6 minutes

Gemini: Google’s largest and most capable AI model

Google finally unveiled their large language model (LLM) called Gemini, and here’s what you need to know:

• Starting now, Bard will use a fine-tuned version of Gemini Pro for more advanced reasoning, planning, understanding and more.

• Starting on December 13, developers and enterprise customers can access Gemini Pro via the Gemini API in Google AI Studio or Google Cloud Vertex AI.

• Gemini Ultra will be available to developers and enterprise customers early next year, same with Bard Advanced.

⚡️Insights:

In the Gemini vs GPT-4 tables above, it looks like Gemini is beating GPT-4 in almost every way. However, do keep in mind that GPT-4 was only trained up to Aug 2022 — basically a 1+ year old model at this point. And yet Gemini isn’t exactly destroying the competition by categories, many of which the model only outperformed GPT-4 by a few percent.

Also, it’s interesting to see Google pulling a fast one on OpenAI after the OpenAI fiesta. Especially after the Gemini event was previously rumored to be delayed.

Elon Musk Releases Generative AI Chatbot Grok on X

xAI, Elon Musk’s AI company, released Grok and will be available to all X Premium+ subscribers ($16/month). The billionaire has described Grok to have a “rebellious streak” that would answer “answer spicy questions that are rejected by most other AI systems,” i.e. giving potentially politically incorrect answers.

Grok was trained with data up to Q3 2023 but it can incorporate real-time data from tweets on X, which technically makes it an AI that’s constantly kept up to date.

Grok has more of an edgy personality than other models, it tells jokes or roasts the user if asked to. However, it certainly is not as adept at performing complicated tasks like other models.

⚡️Insights:

There’s finally a reason for users to get the X Premium+ subscription. Remember when X wanted to sell the blue verified check mark for money? And it backfired and became a meme instead?

At least now the subscription comes with a product that’s different from the rest in the market (AI with a spicy personality), for the users who want to pay for it. This could be another major income stream for X, after Musk dropped the F-bomb on advertisers💥.

EU Agrees ‘Historic’ Deal with World’s First Laws to Regulate AI

The European Parliament and EU member states have agreed on comprehensive laws to regulate artificial intelligence. This puts the EU in the lead in regulating AI compared to the US, China, and the UK.

The specific details of the law are not provided, and it will not be in effect until at least 2025. The underlying principles follow a “human-centric approach respecting fundamental rights, human values, building trust, building consciousness of how we can get the best out of this AI revolution.”

This means that real-time surveillance and biometric technologies like emotional recognition are banned for commercial purposes. And police abuse is mitigated and the tech could only be used in exceptional cases, including unexpected threats of a terrorist attack, the need to search for victims and in the prosecution of serious crime.

🧠Extended reads:

In the United States' context: MIT group releases white papers on governance of AI

Toxic Productivity

Hustle culture and Covid are probably 2 main factors why toxic productivity became more prevalent. The former makes you feel guilty for not be constantly hustling and the latter blurred the line of workplace and rest place for many people.

Here are some methods to mitigate toxic productivity:

1. Recognize and acknowledge your time and energy are finite resources.

2. Set clear boundaries between work and personal life and define specific working hours.

3. Prioritize tasks and learn to delegate or say no when necessary.

4. Embrace breaks and understand it’s more productive to take them.

Meta and IBM launch new group for open-source generative AI development - Link

MIT researchers develop a customized onboarding process that helps a human learn when a model’s advice is trustworthy. - Link

That’s It For This Week, Catch Y’all Same Time Next Week!