AI Wild West Week

Author

Anshuman Pandey
February 21, 2025

Every week in the AI world is new products week - dozens of new tools come out every day. However, this one’s special. So, here’s the wrap of the ones that might be most interesting to you!

Open AI demos a B2B sales bot!

Open AI demo-ed an autonomous sales agent in Tokyo this week, rather quaintly, without the fanfare they’re used to. What can it do?

  • Process inbound leads

  • Identify which ones are worth pursuing based on the lead’s publicly available data, like revenue, user base, etc.

  • Send personalized emails to engage the lead

  • Walk through the process from inquiry to meeting

Well, it’s good news that OpenAI has figured out that sales is the most lucrative use case of them all. If you’d like to see how it works, watch this video demo (starting at 1:07:00).

By the way, the video also demos their latest product Deep Research.

Microsoft launches Muse

OpenAI’s close partner, Microsoft, has gone the other direction with their new product, Muse, a GenAI model for gameplay ideation. It can generate game visuals and controller actions.

If you’re interested in this space, the detailed announcement from Microsoft is here. It also includes some great examples, demos, and discussions on possibilities.

Google presents AI co-scientist

Yes, you read that right. Google is pitching an AI agent as a ‘scientist’ to work with researchers to “accelerate the clock speed of scientific and biomedical discoveries.” Simply put, AI co-scientist is a multi-agent system with reasoning capabilities similar to the scientific method.

While most AI models are accused of regurgitating already-existing information, Google claims that its AI co-scientist is designed to “uncover new, original knowledge and to formulate demonstrably novel research hypotheses and proposals, building upon prior evidence and tailored to specific research objectives.”

With great possibilities come great risks

For starters, outgoing Federal Reserve Vice Chair Micheal Barr suggests, "The very attributes that make genAI attractive — the speed, automaticity and ability to optimize financial strategies — also present risk,” including increasing market volatility and stoking asset bubbles and crashes.”

Insight: Before you think of GenAI, think governance.

Ever asked Google about your headache only to worry that you have some incurable disease? Now, that’s happening with ChatGPT.

“More than half (61%) had asked at least one question that would usually require clinical advice.” — see the problem?

Insight: Even if you’re not in healthcare, replacing processes requiring human expertise with AI is too risky. So, prioritize the expert.

Recently, AI product companies have been positing that giving AI more ‘thinking time’ will make it more efficient and even accurate. Essentially, training AI like a child, “take more time and see if you can do better.”

Forbes columnist Lance Elliot finds that this might not always be fruitful.

Insight: More isn’t always better. In fact, smaller models with thoughtful fine-tuning could be a better approach.

If you’d like to see it in practice, hit reply, and I’ll show you a demo of what we’ve done for our clients.

P.S. DeepSeek R1 is now available on Tune Chat