Dear Reader,
Welcome to the 49th edition of the good reads newsletter by Malpani Ventures. Sharing your weekly dose of articles for this weekend’s reading!
This week we’ve chosen to dive deep into the new, fascinating and exciting world of AI & Generative AI!
The Implications of Today's HUGE AI Announcements
Last week, Google announced Generative AI for Workspace, and OpenAI announced GPT-4. All the generative AI boys and girls have since gone into fanboy and fangirl mode. Some takeaways:
Google workspace
Google Workspace is rolling out Generative AI to its productivity tools, including Gmail, Docs, Sheets, Slides, and Images, among others
The AI features will be available to select testers this month and will have a fuller rollout to Google Workspace customers soon after
The cross-platform and cross-pollination benefits of Google's new AI features in Workspace will supercharge adoption and change the way people think of using these apps
The upcoming AI announcement event from Microsoft suggests that the AI wars are heating up, and innovation will be the ultimate winner
GPT-4
GPT-4 is a significant upgrade from GPT-3.5 and outperforms it on almost every type of exam, with better accuracy, more steerability, and support for visual inputs
GPT-4 has 8x more context than ChatGPT and hallucinates 40% less than GPT-3.5, making it more accurate
GPT-4 is available now in ChatGPT Plus with usage caps and access to the API is waitlisted
GPT-4's sophistication and nuance confirm the trajectory of society has been shifted by AI and raises the question of how long until we get programs that are 10x better than humans at most major tasks. Key takeaway: The rapid advance of AI raises both awe and dread
How Will Generative AI Disrupt Video Platforms?
https://hbr.org/2023/03/how-will-generative-ai-disrupt-video-platforms
Say hello to generative AI - a game-changing technology that can create text, images, audio, and even video with just a prediction of the next word or pixel - making video content creation as easy as a breeze.
This comes at a time when large content players like Netflix, TikTok, and YouTube are already struggling to determine content relevance and engagement, and this can make their problems even worse. Will it create a new value chain for video content creation, or will these companies rethink their approach?
What seems to be the issue? Closed, Hollywood-style content production is expensive and slow. At the same time, open, user-generated content creation requires balancing incentives that can be tricky to achieve - we know the spam & quality issues.
Will we see disruptors like Netflix, TikTok & YouTube disrupt the space again or be disrupted? Time will tell.. but content aficionados are in for a treat irrespective
Chatbots, deepfakes, and voice clones: AI deception for sale
https://www.ftc.gov/business-guidance/blog/2023/03/chatbots-deepfakes-voice-clones-ai-deception-sale
A series and cautionary article from the FTC. The use of AI in creating and spreading fake content is a growing threat that companies need to address. As generative AI becomes more advanced and accessible, scammers can use them to disseminate fake content to large groups or target specific individuals. It asks companies making such tools whether they should even be making them - and/or think about how they mitigate the risks.
Early thoughts on regulating generative AI like ChatGPT
Generative AI models like ChatGPT have gained public attention, but such models still have some limitations. Generative AI systems usually focus on producing aesthetically pleasing imagery, compelling text, or coherent audio outputs. These systems are different from traditional AI systems, which often estimate a specific number or choose between a set of options. Generative AI produces convincing content but does not necessarily comprehend the content it generates. Generative AI is often used for commercial purposes, where it can produce marketing content or generate code. However, the use of generative AI for malicious purposes, like disinformation campaigns and online harassment, is a concern. As generative AI systems become more common, distinct policy interventions may be necessary to address the risks. The policy challenges posed by generative AI are relatively recent, and it's not clear which specific category to focus on.
And before you ask, yes we asked ChatGPT to summarize these posts 😉
Until next time!