ChatGPT’s Canvas, Google’s AI Rivalry, and Microsoft Copilot Upgrades: Key Advancements in AI

AI Innovators!

Welcome to the latest edition of the DSA Newsletter! We’re excited to share groundbreaking advancements in artificial intelligence that are transforming industries and enhancing user experiences. This week, we highlight key developments that showcase the power of AI in various fields.

In today’s AI?Brief:

• ChatGPT’s Canvas: Enhancing Collaborative Writing. How will this new interface revolutionize collaborative projects in writing and coding?
• Google’s Reasoning AI: A New Rival for OpenAI. Can Google’s advanced reasoning capabilities compete with OpenAI’s o1 system?
• Microsoft Copilot: Voice and Vision Upgrades. What do these enhancements mean for user interaction and productivity?
• Google Ads in AI Overviews: A New Approach. Will integrating ads into AI search summaries improve or complicate user experience?
• Liquid AI’s Foundation Models: A Shift in Architecture. What advantages do Liquid Foundation Models offer over traditional architectures?

Join Us!

Stay informed about the latest trends and innovations in AI. Sign up for our newsletter, follow us for updates, and share your insights to connect with our community! What new AI jobs are emerging in this rapidly evolving field? Let’s explore together!

Listen to the DSA Podcast for the latest insights in AI!

ChatGPT gets a collab boost with?Canvas

The Brief: OpenAI just launched Canvas, a new ChatGPT interface release that enables more collaborative writing and coding projects beyond simple chat interactions with new editing features, shortcuts, and added contextual knowledge.

The details:

• Canvas opens in a separate window alongside the chat, allowing users to directly edit and refine specific aspects of an output.
• New features include inline feedback, targeted editing, and shortcuts for tasks like adjusting text length, changing reading levels, or debugging code.
• In tests, using GPT-4o with Canvas led to a 30% accuracy and 16% quality boost compared to using the model without the interface.
• Canvas is rolling out in beta to Plus and Team users, with a broader release expected later.

Why it matters: ChatGPT’s first major UI change takes a leap towards more nuanced, moldable interactions?—?while also inheriting novice-friendly features seen in other rivals with easy-to-use shortcuts. The simple chatbox was a good first step for human-AI interactions, but more power and capabilities require new collaborative processes.

Google developing reasoning AI to rival?OpenAI

The Brief: Google is reportedly making significant strides in developing AI models with advanced reasoning capabilities similar to OpenAI’s o1 system, intensifying the rivalry between the two AI giants.

The details:

• Multiple teams at Google are working on AI that can solve complex, multi-step problems, according to Bloomberg.
• The AI uses chain-of-thought prompting, a technique created by Google, to tackle complex math and programming problems by ‘thinking’ before responding.
• Google is taking a more cautious approach to its releases than OpenAI but has already debuted math-focused reasoning models like AlphaProof and AlphaGeometry 2.
• Microsoft also infused reasoning capabilities into its Copilot assistant this week, leveraging OpenAI’s o1 model.

Why it matters: Human-like reasoning and agentic capabilities are clearly the two major developments on every AI firm’s roadmap, and the release of o1 may have signaled a new phase in the LLM race. The question is?—?will OpenAI’s speed keep it a step ahead, or is the competition for top-tier models about to get a whole lot tougher?

Microsoft Copilot gets voice, vision?upgrade

The Brief: Microsoft just announced a slew of AI upgrades coming to its Copilot assistant for Windows PCs, including new vision and voice capabilities, personalization enhancements, a re-release of the controversial Recall feature, and more.

The details:

• Copilot Voice allows users to interact with natural speech, adding conversational and intuitive communication similar to OpenAI’s Voice Mode.
• Copilot Vision enables the AI to understand and interact with web content a user is viewing, offering context-aware help within the Microsoft Edge browser.
• ‘Think Deeper’ gives Copilot new enhanced reasoning capabilities using chain-of-thought reasoning powered by OpenAI’s o1 model.
• Microsoft’s ‘Recall’ feature is set to return, requiring an opt-in with upgraded privacy and security measures.
• Microsoft AI CEO Mustafa Suleyman highlighted Copilot’s ability to ultimately ‘act on your behalf’ and adapt to user’s personal preferences and needs.

Why it matters: Microsoft is bringing the heat with these major Copilot upgrades, levelling up the assistant to align with the latest cutting-edge AI features across the industry?—?while bringing users one step closer to a truly agentic experience.

Google rolls out ads in AI Overviews

The Brief: Google just announced the introduction of ads to its AI Overview search summaries and the launch of several new AI-powered search capabilities, such as video understanding and voice input.

The details:

• Ads will now appear within and alongside AI Overviews for ‘relevant queries’ on searches in the United States.
• The redesigned AI Overview format will now add prominent in-text links to better source websites for the curated information.
• New AI-organized search results pages are rolling out that surface relevant, more diverse content?—?starting with recipe and meal inspiration queries.
• Google Lens is getting video understanding capabilities and voice input options for visual searches.
• The Android ‘Circle to Search’ feature also lets users identify songs playing in videos or streaming content.

Why it matters: Google’s first AI Overview experience didn’t exactly go as planned. However, with heavy competition from Perplexity and chatbot rivals, Google’s search future clearly has AI at its core, regardless of the bumps along the way. But infusing paid ads into AI Overviews could be a slippery slope?—?will Gemini be next?

Liquid AI unveils efficient new LFM?models

The Brief: Liquid AI just introduced a new series of AI models called Liquid Foundation Models (LFMs), challenging the traditional transformer architecture while achieving state-of-the-art performance and enhanced memory efficiency at smaller model sizes.

The details:

• The company released its LFMs in 1.3B, 3B, and 40B parameter sizes, based on a new architecture utilizing computational units rooted in dynamical systems rather than traditional transformers.
• The models surpass transformer-based counterparts like Meta’s Llama 3.2 and Microsoft’s Phi-3.5 on major benchmarks like MMLU.
• LFMs require significantly less memory for inference, particularly with long-context tasks?—?supporting up to 32k tokens while maintaining memory efficiency.
• The models are not open-source and are only currently available via the company’s Lambda (Chat UI and API) and on Perplexity AI.

Why it matters: Liquid AI’s LFMs are a significant shakeup from the transformer architecture standard that has dominated models since 2017. The benchmarks show that there is more than one formula for achieving state-of-the-art AI performance?—?and could open new possibilities for more efficient and accessible AI systems.

OpenAI secures SoftBank funding as Apple exits?raise

The Brief: Despite Apple reportedly no longer participating in OpenAI’s upcoming funding round, the AI giant has secured billions of dollars from Japanese investment giant Softbank, Microsoft, and Thrive Capital.

The details:

• OpenAI is rumored to be raising up to $6.5B via convertible notes, at an eye-popping $150B valuation.
• Microsoft plans to participate with an additional $1B, adding to its previous $13B investment in the AI giant.
• Investment firm Thrive Capital is also investing $1B, with a reported option to add an additional $1B the following year based on revenue goals.
• The Wall Street Journal reported that Apple is no longer involved in the funding round, despite partnerships with OpenAI and its inclusion in Apple Intelligence.
• The raise comes amid OpenAI’s controversial restructuring to a for-profit entity, with Sam Altman denying rumors that he will receive equity in the move.

Why it matters: OpenAI’s latest raise and for-profit turn is another saga in its convoluted and controversial business structure. Despite the recent high-profile departures and continued drama, the ChatGPT maker is still clearly seen as a top horse to bet on in the AI boom?—?and there is no shortage of major players who want in.

AI TRAINING

Unlock multiple ChatGPT tools in one?chat

The Brief: ChatGPT’s new shortcut feature lets you instantly switch between image generation, web search, and advanced reasoning tools directly in one chat?—?avoiding the need to reset chats.

Step-by-step:

1. Start a new chat in ChatGPT and type “/” in the input field.
2. Choose from three options: Picture (DALL-E), Search (web), or Reason (GPT-o1).
3. For images, use “/picture [description]” (e.g., “/picture quantum computer”).
4. For web searches, use “/search [query]” (e.g., “/search quantum computer”).
5. For complex reasoning, use “/reason [task]” (e.g., “/reason Explain quantum computing”).

Pro tip: When using the /search command, try adding “latest” or a specific year to your prompt.

….

Turn YouTube videos into AI-powered podcasts

The Brief: NotebookLM’s latest update allows users to transform lengthy YouTube videos into concise AI-generated podcasts, saving time and enhancing study efficiency.

Step-by-step:

1. Visit NotebookLM and create a new notebook.
2. Click on “Link” in the source selection area, choose “YouTube” and paste your desired YouTube video URL.
3. Select “Generate” in the Audio Overview section to create your AI podcast.
4. Interact with your podcast by playing it, asking questions via chat, or generating additional study materials.

Pro tip: Use the chat feature to ask specific questions about the content, turning your AI podcast into an interactive study session!

…

Automate video analysis with Gemini?AI

The Brief: Google Gemini on AI Studio can analyze videos and provide transcripts, tags, subtitles, and translations to simplify and speed up your content creation workflow.

Step-by-step:

1. Access Google Gemini on AI Studio and select “Gemini 1.5 Pro 002” from the Models menu.
2. Upload your video and use this prompt: “Analyze this video and provide the transcript, 5 title ideas, and categorized tags.”
3. Follow up for improvements: “Suggest 5 content improvements, 3 promo clip ideas with timestamps, reach expansion tips.”
4. Implement insights to optimize SEO, create promo clips, and expand your audience reach through translation.

Pro tip: Regularly analyze your video content with Gemini to track improvements and identify trends in your content over time.

AI RESEARCH

MIT’s ‘Future You’ taps AI to speak with older?self

The Brief: Researchers at MIT have developed an AI system called “Future You” that allows users to interact with and ask questions to a simulated version of their older selves.

The details:

• The system uses personal information provided by users to create a realistic future self-simulation, including generating an age-progressed photo.
• Users engage in text-based conversation with an AI-generated 60-year-old version of themselves, capable of answering questions and offering insights.
• In a study of 344 participants, those who used Future You reported decreased negative emotions and anxiety.

Why it matters: While aging simulation apps are constantly going viral, the implications of AI-driven psychological support are massive. With AI’s ability to create and simulate highly personalized, empathetic experiences, studies like Future You are only scratching the surface of the future of therapy and psychology.

…

Black Forest Labs unveils Flux 1.1?Pro

The Brief: Black Forest Labs just released Flux 1.1 Pro, a significantly upgraded version of the startup’s text-to-image AI model, and a new API for developers.

The details:

• Flux 1.1 Pro generates images six times faster than Flux 1 Pro while improving quality and prompt output adherence.
• The model tops the Artificial Analysis image arena leaderboard against rivals like Midjourney, Ideogram, and DALL-E, tested under the codename ‘blueberry.’
• 1.1 Pro will be a paid model available through partners like Together AI, Replicate, FAL AI, and Freepik, unlike the open-source Flux 1 that powers xAI’s Grok.
• BFL’s API allows third parties to integrate the model into their apps, and the 1.1 Pro model costs?.05c / image.

Why it matters: From OpenAI’s strawberry to BFL’s blueberry, fruit codenames are having a moment! 1.1 Pro looks to raise the already incredibly high text-to-image bar, continuing to push the boundaries of realism and image generation quality?—?now equipped with a turbocharged speed increase as well.

QUICK HITS

OpenAI’s Sora research lead Tim Brooks announced on X that he is leaving the company to join Google DeepMind, where he will work on ‘video generation and world simulators.’

Fourier launched GR-2, the company’s second-generation humanoid robot, which features improvements to battery life, hand dexterity, mobility, and a new developer kit.

OpenAI secured a new $4B credit facility from major banks, boosting its total liquidity to over $10B to fuel future growth and innovation.

AI Coding startup Poolside announced a $500M Series B funding round to accelerate progress towards AGI, bringing the company’s valuation to $3B.

Google agreed to invest $1B into Thailand to expand AI and cloud infrastructure in Southeast Asia, aiming to build new data centers amid increasing regional competition.

TikTok parent company ByteDance is reportedly planning to develop a new AI model primarily using Huawei chips, diversifying from U.S. suppliers like Nvidia to counteract export restrictions.

Artisan AI secured $7.3M in seed funding for its sales-focused AI virtual employees, with its first AI assistant Ava already assisting over 120 companies on the platform.

Luma Labs upgraded its Dream Machine AI video model speed, allowing for full-quality generations in under 20 seconds.

Qodo announced a $40M funding round for its AI-powered code testing software, with plans to expand services and target larger enterprise clients.

The U.S. Commerce Department unveiled a plan to award $100M for AI semiconductor research, hoping to spur the development of more sustainable materials.

AI reading coach startup Ello launched ‘Storytime’, a new feature allowing kids to create personalized stories using AI.

Google released Gemini 1.5 Flash 8B, a lightweight, cost-effective variation with a 50% cost reduction and 2x higher rate limits than 1.5 Flash.

SPONSOR US

Get your product in front of over 650k+ AI enthusiasts

Our newsletter is read by thousands of tech executives, investors, engineers, managers, and business owners around the world. Get in touch today.

How would you rate today’s newsletter?

Comment below to help us improve the newsletter for you.