Hello Again!

I hope AI Emergence is capturing your interest – it's designed to bring you insights covering AI while it is Emerging :)

If you are new here, we are building Human Intelligence about Artificial Intelligence while AI is Emerging!

What would be the format? Every week, we will break the newsletter into the following sections:

• The Input - All about recent developments in AI

• The Tools - Interesting finds and launches

• The Algorithm - Resources for learning

• The Output - Our reflection 

• Question to ponder before we meet next!

Table of Contents

• The Input

  • OpenAI Trying to Level Field this Election Season …

  • 2024: The Dawn of Humanoid Robotics?

  • Copilot Pro: Microsoft's New Leap in AI Productivi …

  • Google's InseRF: Pioneering Text-to-3D Technology

  • Samsung Galaxy S24 Series: AI Takes Center Stage

  • HCLTech and SAP's Partnership for Generative AI

• The Tools

• The Algorithm

• The Output

  • Question to ponder before we meet next!

The Input - Recent Developments in AI

OpenAI Trying to Level Field this Election Season in the USA

As the 2024 US elections commence, Generative AI emerges as a key tool for politicians. However, it also brings challenges like deceptive content and Deepfakes. OpenAI has outlined its strategy to combat these issues and ensure election integrity: 

• Preventing AI Tool Misuse: OpenAI restricts its tools from being used for political campaigning or creating misleading chatbots, aiming to maintain integrity and trust. Applications that distort voting processes or hinder democratic participation are also prohibited, with provisions for users to report any violations in new GPT models.

• Enhancing Transparency in AI Content: OpenAI is adopting cryptographic methods to track the origin of images created by DALL-E 3. This move will improve the detection of AI-generated content and help voters evaluate the reliability of such information.

• Promoting Accurate Voting Information: Through ChatGPT, OpenAI plans to direct users to authoritative voting resources like CanIVote.org for US election queries. This initiative is designed to provide voters with precise and reliable information, supporting informed decision-making.

This election period is likely to be a turning point, showcasing significant advancements in Generative AI and setting new precedents for future global elections. (source)

Overall- I think it is a good step from OpenAI to mitigate the risk - but I think the far bigger risk in elections could be spreading agenda not directly related to elections - which is still unchecked. Of course - that is a far more difficult problem to solve. But with elections for a significant part of the world (including the US and India) coming up in the next 12 months - this is going to play a key role and will be evolving significantly.

2024: The Dawn of Humanoid Robotics?

The AI robotics field is bustling with exciting developments!

Elon Musk's recent reveal on his X account features Tesla's Optimus robot impressively folding a t-shirt. The video highlights the robot's dexterity, showcasing its human-like ability to handle a garment with ease.

Meanwhile, 1X Technologies, a Norwegian robotics startup and a beneficiary of OpenAI's backing has just secured a hefty $100 million in a Series B funding round. This round saw contributions from several major players, including Samsung.

In another advancement, LimX's humanoid robot demonstrates its capability to navigate stairs autonomously. In trials, it successfully identified and adapted to stairs, slopes, and curbs on the fly, showcasing its advanced mobility.

I think that these are still early days for Robotics - but its impact would be huge. In a few years, think of robots at your house, in your office, and other places taking care of all the mundane tasks and helping you out!
(sources- 1, 2, 3)

Copilot Pro: Microsoft's New Leap in AI Productivity

Microsoft introduces Copilot Pro for Microsoft 365, offering advanced AI features for creative and professional tasks. With its integrated Image Creator tool, users get enhanced performance in applications like Word and Excel, across multiple devices. (source)

If you are not a GPT4 user, this is a great addition where it matters the most. With the little time I had playing with these, the addition looked impressive in Word, PowerPoint, Excel, and Outlook!
I look forward to what it does to the life of Management Consultants!

Google's InseRF: Pioneering Text-to-3D Technology

Google unveils InseRF, a groundbreaking method for adding objects into 3D scenes via text descriptions. It blends objects realistically into 3D environments, using innovative single-view reconstruction for seamless 2D to 3D transitions. (source)

Samsung Galaxy S24 Series: AI Takes Center Stage

Samsung's latest Galaxy S24 series was unveiled yesterday, with a deep focus on artificial intelligence. The event was positioned as unveiling “Galaxy AI”.

The Galaxy S24 Ultra, powered by Qualcomm's Snapdragon 8 Gen 3 chipset, introduces new AI capabilities. Notable features include item recognition through the camera, AI-enhanced photo editing, and real-time speech translation.

Samsung also addresses privacy concerns; AI-edited photos will include metadata logs of alterations and a watermark to indicate AI usage, balancing innovation with user privacy and copyright considerations. 

It looks interesting - no major surprises on the AI front except one that Samsung went ahead with Google Gemini models! (source)

HCLTech and SAP's Partnership for Generative AI

HCL Technologies collaborates with SAP to drive Generative AI in business, establishing a GenAI Center of Excellence. This alliance will integrate advanced AI capabilities into SAP’s cloud solutions, targeting transformative changes across industries. (source)

The Tools-Interesting Finds and Launches

• Personalized AI with Microsoft Copilot GPTs: Microsoft is enhancing user experience with customizable Copilot GPTs, catering to interests like fitness, travel, and cooking. The upcoming Copilot GPT Builder tool for Copilot Pro users will enable the creation of tailor-made AI assistants. (source)

• Sarvam AI's OpenHathi LLM: A Hindi Language Marvel: Indian startup Sarvam AI debuts OpenHathi LLM(OpenHathi-Hi-v0.1), a Hindi-specific language model built on Meta AI’s Llama2-7B architecture. This model promises performance rivaling GPT-3.5 for Indic languages. (source)

• Amazon's AI Tool for Enhanced Product Queries: Amazon is testing a new AI feature to simplify customer inquiries about products. This innovation aims to reduce the need for browsing through reviews and listings, as part of Amazon's expanding AI applications in customer interaction. (source)

The Algorithm - Resources for Learning

• Chip Huyen's article sheds light on the probabilistic nature of AI models, emphasizing their creative yet inconsistent outputs and the significance of test time sampling. (source)

• This novel paper presents a simpler alternative to RLHF for aligning language models with human preferences, focusing on the training efficiency of LLMs. (source)

• In my recent episode of 'Leading with Data', Dr. Gautam Shroff talks about the Neuro-symbolic approach to AI in his work. (source)

• AI and next-gen cloud tech are supercharging scientific discovery, especially in chemistry and materials science, through a Microsoft-PNNL partnership. (source)

• Google Deepmind releases AlphaGeometry, an Olympiad-level AI system for Geography. This reinforces the discussion I had with Dr. Shroff about Neuro-symbolic approaches to improving the problem-solving abilities of AI. (Source)

• In "Unconfuse Me with Bill Gates," Sam Altman talks about AI's rapid integration into society and the imperative for ethical guidelines and global oversight. What stands out is Sam saying that these would be the stupidest models we are going to see in the coming years! (Source)

The Output

A week full of promises, actions, new findings, and launches!
I think the question about AI’s impact on elections is going to be a recurring theme for the year and something we all will be playing a catch-up game on.

Question to ponder before we meet next!

Last week I asked your views on AI devices - There have been a few “AI device” launches that are positioning themselves as AI phones - Humane, Rabbit r1, and a few more. Do you think that AI phones will replace smartphones in the 3-5 years time frame? Roughly 66% of you think that AI phones will replace smartphones in this time frame.

I think otherwise. I think that these devices would complement existing smartphones and not replace them. I have tried not using smartphones in the past for using different ways/strategies - no phone at all, no WhatsApp, no communication apps, with human assistance - but all trials ended up with me frustrated about the overheads. I feel smartphones will be needed to regularly consume content, and stay connected, and over time smartphones will add an AI interface.