Microsoft is making OpenAI's o1 free
Nvidia CEO: everyone should get an AI tutor
Reasoning AI why now
The #1 AI Meeting Assistant
Typing manual meeting notes drains your energy. Let AI handle the tedious work so you can focus on the important stuff.
Fellow is the AI meeting assistant that:
✔️ Auto-joins your Zoom, Google Meet, and Teams calls to take notes for you.
✔️ Tracks action items and decisions so nothing falls through the cracks.
✔️ Answers questions about meetings and searches through your transcripts, like ChatGPT
Try Fellow today and get unlimited AI meeting notes for 90 days.
***
Microsoft is making OpenAI's o1 free
Microsoft is making a bold move to democratize high-end AI reasoning. Offering free access to OpenAI’s powerful o1 model through Copilot’s newly introduced “Think Deeper” toggle. The key detail here is “free.” OpenAI debuted the o1 model in December, touting it as the most advanced AI model available. However, access came at a steep price—requiring a ChatGPT Pro subscription costing $200 per month for unlimited use.
That changed on Wednesday when Mustafa Suleyman, Microsoft’s chief of AI, announced that o1 would be accessible to all Copilot users “everywhere at no cost.” The model will be available through Copilot’s “Think Deeper” function, which takes a few extra seconds to generate more comprehensive, well-researched responses.
Users can access the feature via the Windows app or directly through copilot.microsoft.com. A Microsoft account login is required. The “Think Deeper” feature functions as a simple toggle—users just need to switch it on before submitting a query. Designed for more in-depth analysis, Think Deeper enhances Copilot’s ability to provide detailed, well-reasoned answers, making it a valuable tool for users seeking deeper insights....
***
Get Your Team Booked on 3.8 Million Podcasts Automatically
It's 2025. Want to finally be a regular podcast guest in your industry? PodPitch will make it happen. Even the beehiiv team uses it!
The best way to advertise isn't Meta or Google – it's appearing on podcasts your customers love.
PodPitch.com automates thousands of weekly emails for you, pitching your team as ideal guests.
Big brands like Feastables use PodPitch.com instead of expensive PR agencies.
Get your FREE personalized demo!
***
Nvidia CEO: everyone should get an AI tutor
Nvidia CEO Jensen Huang believes humans have a lot to learn from artificial intelligence—and he’s already using it to stay ahead. As AI innovation accelerates, Huang advises those looking to keep pace with the future to embrace the digital educators available at their fingertips, including OpenAI’s ChatGPT and xAI’s Grok.
“The knowledge of almost any particular field, the barriers to that understanding, have been reduced. I have a personal tutor with me all the time,” Huang said in a recent interview. He emphasized that AI can teach a broad range of subjects, assist with programming, and help users write, analyze, think, and reason.
“All of those things are really going to make you feel empowered, and I think that’s going to be our future,” he added. Huang described AI as a tool that boosts confidence, allowing individuals to take on increasingly ambitious challenges.
“It actually empowers me and gives me the confidence to go tackle more and more ambitious things,” he said. “It’s going to empower you; it’s going to make you feel confident. I feel more empowered today, more confident to learn something today.”….
***
Reasoning AI why now
There is a flurry of activity in the world of AI, as hundreds of startups scramble to build the next big thing. In recent weeks, we've seen a string of reasoning models emerge, including OpenAI's o1 and o3, Google DeepMind's Gemini 2.0 Flash Thinking, and now R1. What's driving this sudden proliferation? According to experts, the answer lies in the maturity of the underlying base models.
"What R1 shows is that with a strong enough base model, reinforcement learning is sufficient to elicit reasoning from a language model without any human supervision," explains Lewis Tunstall, a scientist at Hugging Face. In other words, the top US firms may have cracked the code, but they were keeping the process under wraps. "It seems that there's a clever way of taking your base model, your pre-trained model, and turning it into a much more capable reasoning model," says Zeiller.
The game-changer, however, is that DeepSeek has now published how they achieved this feat. "And it turns out that it's not that expensive a process," says Zeiler. "The hard part is getting that pre-trained model in the first place." As Andrej Karpathy revealed at Microsoft Build last year, pretraining a model is the most time-consuming and costly part of the process.
If building reasoning models is not as difficult as previously thought, we can expect a proliferation of free models that are far more capable than we've seen before. With the know-how out in the open, Friedman believes this could be a "monumental moment" for collaboration between small companies, potentially blunting the edge that the biggest players have enjoyed….