Microsoft And OpenAI Are Teaming Up

01-24-2023

⚡ Today’s Highlights

  • 🚨Microsoft and OpenAI VS The World

  • 🤖The Singularity inches ever closer

  • 👃TurnItIn is learning to sniff out AI

📰 Today's Top Stories

(4 min read) (Source: Microsoft)

Microsoft is making some tsunami-sized waves in the AI world this week by announcing that it's going to be extending its partnership with OpenAI in a multiyear, multibillion-dollar deal. The company released no official details but a source close to Microsoft reported to the New York Times that the tech giant is investing $10 billion dollars in the San Francisco-based lab, which already has more than $3 billion from Microsoft in the bank.

With this investment, the power couple seems to be solidifying their positions as leaders in the AI space. They hope to push the boundaries of AI with a specific focus on generative AI and crush the competition in the process. OpenAI, which was created as a non-profit in 2015 but converted to a for-profit company in 2018 to fund its research, has already made waves with its A.I. system GPT-3.

Microsoft is already integrating GPT-3, DALL-E, and other OpenAI technologies in products like GitHub. Both companies have stated that their ultimate goal is to develop A.G.I., an artificial general intelligence that can do anything the human brain can do.

(5 min read) (Source: Popular Mechanics)

The term "singularity" in AI refers to the point when artificial intelligence surpasses human control and transforms society (and maybe starts building time traveling robots). One way to measure progress towards singularity is through an AI's ability to perform tasks comparable to human skills, such as speech translation.

Translated, a Rome-based translation company, has been tracking its AI's performance using a metric called "Time to Edit", which calculates the time it takes for professional human editors to fix AI-generated translations compared to human ones. Over an 8-year period, Translated's AI has shown slow but steady improvement, and is predicted to be as good as human-produced translation by the end of the decade or even sooner.

While this is a cool way to quantify the singularity, it's one of many definitions and the true "technological singularity" remains elusive. It's still worth mentioning that we're getting closer to the point where the machines will be able to translate as well as humans, and before you know it, they'll be planning our demise in over 7000 different languages!

(10 min read) (Source: The Guardian)

We've all seen the kinds of "art" that AI is capable of. Some are beautiful, some are horrifying, some are anime renditions of our faces that make us feel better about ourselves. Thanks to AI image generation companies like MidJourney, Stable Diffusion, and Deep Dream Generator, the internet is being populated with images created by AI that many think are just mashups of existing art rather than original work. These companies allow users to create images from text prompts for free, but some experts have raised concerns about the implications for the art industry and the potential to devalue the skill of illustration.

Author and illustrator Rob Biddulph argues that true art is about the creative process, not just the final product, and that pressing a button to generate an image is not a creative process. Meanwhile, the #NotoAIArt campaign is raising concerns about the legality of AI image generators and the potential for them to rely on copyrighted images.

While some artists advocate for license-based models to protect intellectual property, others believe that AI image generation may impact smaller jobs that up-and-coming artists rely on to build portfolios. The silver lining? Many illustrators believe that kids and young people will not find AI art to be convincing.

(8 min read) (Source: The Register)

Turnitin, the company known for stopping plagiarists dead in their tracks, is taking on a new role: snitching on AI text generators. With the increasing use of large language models like OpenAI's GPT-3 and ChatGPT in academic settings, concerns about cheating have risen. While some educators see the technology as a way to enhance writing skills, others view it as cheating. Education departments are currently unsure of how to regulate the use of AI text generators.

But in the shadows of the Dropbox, Turnitin has secretly been working on an AI text detection software and it's ready for whatever the crafty students of today can throw at it. It will provide basic AI-generated text detection at first, but the company has plans for more actionable updates for educators in the near future. The tool is based on the perceived difference between human and AI writing.

Currently, AI text generators write by predicting what word should go where in the current context, but they doesn't fully understand the natural ways in which humans communicate. For this reason, AI-generated text can be detected by machines due to patterns in the way it is written.

The software is not intended to ban AI text generators from academic settings, but rather to build a level of trust between students, educators, and our good old pal ChatGPT.

(17 min read) (Source: Forbes)

ChatGPT is quickly establishing that it has the ability to help people navigate the legal system for minor cases in ways that might be cheaper than hiring a lawyer. AI and blockchain technology have the potential to revolutionize the legal industry by streamlining contract review, automating document discovery, and addressing traditional hurdles of privacy, traceability, trust, and security.

Corporate legal teams may see it as a valuable tool for "first draft" tasks, but most law firms will see it as a threat to their business model. Some may use scare tactics to discourage clients from using these tools, but this short-sighted view will likely change as clients exert more pressure for change and competition enters the legal marketplace.

Generative AI tools such as ChatGPT could also have a wide range of uses in areas such as marketing, sales, IT/engineering, risk and legal, HR, and customer service. With accessible and affordable AI tools like ChatGPT, the delivery of legal services could become democratized, potentially transforming the role of lawyers and the judicial system as a whole.

🌎 Popular Content

1. It looks like AI is actually getting pretty good grades! Check out this discussion about ChatGPT passing an MBA exam at Wharton. (Reddit)

2. Flawless AI is doing some amazing things in the movie business, they can visually alter actors to make it look like they’re speaking different languages. (Reddit)

3. Speaking of AI changing up the legal system, the first AI “robot” lawyer is set to represent a defendant in court next month. (Reddit)

4. I feel like they were thinking more along the lines of time traveling assassin than pamper you at the dinner table when they were making this one. (Twitter)

5. This is what happens when you let AI get weird. AI is asking Ben Winshaw if he can let it hold $30…smh. (Vulture)

👀 More Reading

  1. "Deep Learning Market Report - Analysis of growth strategies, marketing channels, and competitors in the industry. Helps businesses expand and understand the market.” (Digital Journal)

  2. Are you ready to trust AI with your insurance information and your insurance rates? Do you trust it to tell you how to respond in emergencies and disasters? (Insurance Journal)

  3. Here are some more insights about how teachers are trying to respond to GPT in the classroom. (Faculty Focus)

  4. Seems like medical advancements are going to accelerate in the next decade. Two studies, published in the journal Nature Medicine, demonstrated that by combining data from wearable technology with advanced medical AI, researchers can identify movement patterns, predict potential diseases, and make clinical trials more efficient. The team behind the studies is made up of AI experts and clinicians. Clinicians + Biotech + AI…here we go.

  5. Suzy, the connected market research cloud, uses GPT to tell us how much we like/hate GPT. Suzy’s survey revealed that:

    1. 35% of consumers are aware of AI tools and most familiar with Grammarly, ChatGPT, and DALL-E.

    2. 37% of AI users would likely purchase an AI tool like ChatGPT.

    3. 78% of AI users would be upset if they were no longer able to use an AI tool.

Powered by AI. Curated and edited by Humans.

How'd you like today's email?

Login or Subscribe to participate in polls.

We'd love to hear from you! Leave us comments or feedback by replying to this email!