Welcome to AI Future - your weekly one-stop shop for all things AI.

The last few years have been dominated by hype. The two kings? Crypto…and now AI. Now, an effort that combines the two. How? Read on…

In the news this week:

• 👁️ Worldcoin launches

• 🕵️‍♂️ Detecting AI is hard

• ❤️ AI detects heart attacks

• 🎒 Law schools split on AI

📰 AI in the News

Worldcoin Wants Your Eyeballs

When Worldcoin launched in 2019, venture capital largely regarded it as a crypto play.

How it works: scan your iris, get some cryptocurrency. Simple enough, and in the backdrop of the crypto bull run of 2021, made complete sense.

Then, crypto crashed, and so did Worldcoin’s relevance and chances of finding funding among Silicon Valley kingmakers.

But as AI took lead on attention-grabbing headlines, Worldcoin has reinvented itself as a solution for a “post-AI” world. That storyline seems to be working, raising $115 million earlier this year.

Here’s the new pitch: by scanning your iris, Worldcoin gives you a digital ID that affirms that you are indeed human. That’s useful in a world where you can’t tell if it’s a human or an AI calling you on the phone.

The kicker: OpenAI CEO Sam Altman is a cofounder. You gotta give it to him. Betting on this in 2019 is impressive second-order thinking.

Worldcoin officially launched last week, having completed scans in major cities like SF, NYC, London, Tokyo and more. So far, about 2-3 million people have had their eyes scanned.

The solution to AI might just be crypto. Maybe there was something behind this hype all along.

Detecting AI-Written Text is Hard

No doubt the most intense conversation re: AI so far has been in education. Put simply, nobody knows how to deal with the flood of ChatGPT in the classroom.

For a short while, it seemed that there would produce tools that could detect AI-generated text. “Academic integrity” company TurnItIn and startups like GPTZero quickly released their attempts. Even OpenAI released an early version of a detector.

Here’s the problem: none of them worked. In fact, they likely caused more troubles than they fixed. Social media flooded with students being falsely accused of using AI text generators like ChatGPT to do their work.

Last week, OpenAI all but admitted it was an unsolved problem. They quietly pulled their detection tool from their market.

For the time being, with AI models proliferating, it seems we won’t know who actually wrote that article you’re reading.

More Stories

• Advice for companies and managers reluctant to embrace AI at work. (link)

• News Corp, The New York Times and Axel Springer all want their cut from AI models, to the tune of billions. (link)

• Spotify wants AI to create more personalized experiences, summarize podcasts, and generate ads in its app. (link)

• Researchers developed a model called "STEVE-1" that can play Minecraft. (link)

• Shopify is launching an AI assistant named "Sidekick" to help business owners make decisions. (link)

• Stability AI releases a new AI image maker called SDXL 1.0. (link)

• OpenAI, Anthropic, Google, DeepMind, and Microsoft formed a group to promote safe and responsible AI systems. (link)

• Amazon is expanding Bedrock, its AI platform, to include conversational agents that can perform tasks. (link)

💡 Industry Insights

🏥 Healthcare

Google/Amazon: Both are looking to healthcare as their first industry to tackle with generative AI. We heard about Google’s MedPaLM being used in clinical settings a couple weeks back. This week, Amazon launched healthcare services via AWS. (link)

Keeping you posted: AI startup OpenEvidence wants medical professionals to have access to chatbots that can reference the latest research vs. ChatGPT, whose data ends in September 2021. (link)

Healthy hearts: AI might be able to detect heart attack 3-10 years before symptoms begin. Fountain Life has developed a solution that can detect uncalcified plaque, which humans can’t see. (link)

👩‍⚖️ Legal

T&Cs: Looking to bring an AI chatbot like ChatGPT to your workplace? Here’s what you need to look for: no training on your data, minimal data retention, IP ownership, IP indemnity. (link)

Application season: While the University of Michigan is banning the use of AI in law school applications, Arizona State says “go ahead”. (Michigan, Arizona State)

📈 Venture Deals

• o9 Solutions, an AI software provider helping global businesses make decisions, raised $116 million. (link)

• Neura Robotics raised $55 million for its cognitive robots that collaborate with people. (link)

• Unstructured helps enterprises organize their data for LLMs and raised $25 million. (link)

• GGWP, an AI platform that moderates behavior in online games, raised $10 million. (link)

• Graft raises $10 million to build an AI development platform for the masses. (link)

🛠️ Latest AI Tools

• Guidde AI is a GPT-powered tool that creates video documentation 11x faster. (link)

• Fillout generates forms for you using AI. (link)

• Rewind AI launches an iOS app to help you search everything you’ve seen. (link)

• Universe is an AI-driven way to build websites. (link)

• TikTok is going crazy for Remini, which can turn any photo into a professional headshot. (link)

• Relay by AngelList uses AI to analyze your investment portfolio. (link)

Share & Subscribe

Did someone send you this newsletter? Subscribe Here.

Know someone who’s interested in learning about AI? Copy and paste this link to others: https://tiferes.beehiiv.com/subscribe