Snap infuses a personality in its new chatbot

At times it feels like AI progress is happening too fast and we need to slowdown and look back how we can maximise the positive outcomes and minimise the negative outcomes.

With that, lets get rolling - 

Here’s what we have lined up for you today :

1. Snap infuses a personality in its new chatbot

2. Windows 11 might as well be called Windows GPT

3. Understanding the role of regulation for AI technologies

Snap infuses a personality in its new chatbot

Yesterday we shared that Snap is launching an in-app ChatGPT like experience for users on their platform powered by the same technology that powers ChatGPT. Today we have additional details about how the company plans to implement this in their product offering. 

Evan Speigel exclaimed that "The big idea is that in addition to talking to our friends and family every day, we're going to talk to AI every day." 

Snap is calling their chatbot My AI which has been purpose built for its young millennial/ GenZ audience. Hence My AI has been trained to have a distinct personality which Snap claims aligns with core values of friendship, learning and fun. 

My AI is only available to Snap+ customers for $3.99/month which opens up an additional revenue stream for the company which has flirted with many different business models over the years. Snap has also explicitly stated that they are purposely being restrictive with what My AI is capable of having seen how other chatbots have responded when being used openly by the general public.

Windows 11 might as well be called Windows GPT

Microsoft is pushing hard for AI to be in every aspect of their operating system stack. The new Windows 11 update will have the Bing taskbar pre-installed. Next up we are most definitely going to see Microsoft deploy its Prometheus model (based on GPT 4) across Word, Powerpoint, Outlook and any other Windows application that you can possibly think of !

Understanding the role of regulation for AI technologies

OpenAI, BBC, Bumble, TikTok are amongst a group of 10 companies that have signed a set of guidelines on how to build, create and share synthetic data/ content responsibly. These voluntary guidelines were put together by Partnership on AI (PAI), an AI research nonprofit. The framework includes a number of recommendations the technology companies and distributors should consider as they increase the deployment of synthetic media created by AI.

The framework specifically focuses on the end users of these technologies and how to best protect them from any hurtful content. It further highlights how consumers need to be made aware of how they are interacting with AIs and the limitations/drawbacks of such systems. There is a fine line between regulation and innovation, one that needs to be balanced to ensure progress. 

Product Watch Verb - 

As AI written books flood the market, this product is tailor built for serious writers who want to enhance their storytelling with creative tools powered by the very technology that “threatens”  to replace them. 

Around the industry - 

1. Using Large Language Models Effectively. What 10 Cutting-Edge Companies Have Learned From Building on Top of LLMs

2. How anybody can use Midjourney as an interior design tool 

3. Microsoft's Multimodal LLM can understand images, text, images with text, OCR, image captioning, visual QA and even solve IQ tests

4. A super effective walkthrough of how ChatGPT is working under the hood !

Time for our weekly Haiku

This time it's about AI regulation 

" Code without limit,

AI's power must be tamed,

Regulation's blade."