🤖 Meta's Game-Changing AI Unveiled!

What’s in today’s Business Edge update?

Meta shakes up the AI world with ImageBind, its groundbreaking multimodal AI model now open-sourced, indicating Meta’s AI strategy plans.

In other news, Airbnb's Q1 revenue beat dampened by booking miss, causing shares to plummet 9%. Meanwhile, Meta revamps its earnings model to reward creators based on Reels' performance, while big tech mergers face increased scrutiny from regulators across the globe.

Let’s get started!

From our partner:

🌐 Big Tech digest

Concise summaries of the most interesting tech news stories from Big Tech.

[1] Meta Unveils ImageBind, a Groundbreaking Multimodal AI Model, Now Open-Sourced

Meta Platforms Inc. has open-sourced ImageBind, an innovative artificial intelligence (AI) model capable of processing six different types of data, potentially paving the way for novel AI applications. The model, which Meta claims outperforms conventional models that process only one data type, can handle images, text, audio, infrared sensor data, depth maps, and data from inertial measurement units (IMUs).

More:

• ImageBind is part of Meta's initiative to develop multimodal AI systems that learn from various data types.

• Unlike typical multimodal models, ImageBind stores multiple data types in a single embedding.

• ImageBind is based on self-supervised learning, which simplifies the creation of training datasets.

• In internal tests, ImageBind excelled in audio and depth data classification tasks, and set a performance record for certain "emergent zero-shot recognition" tasks.

• ImageBind's architecture allows it to analyze several different data types simultaneously.

_{Tags: #Meta, #AI, #Open-source}

Read more about it at —> (read time: 7 minutes)

Insight:

We believe this development represents an industry-wide impact event. Meta's ImageBind opens new frontiers in AI research by handling multiple data types in a unified manner. Its potential to outperform traditional models could significantly reshape how AI systems are designed and used. Furthermore, the open-sourcing of ImageBind means that researchers worldwide can contribute to its development and application, accelerating AI advancement. The model's self-supervised learning approach simplifies training data preparation, which could democratize access to advanced AI capabilities for smaller companies and startups.

The news about ImageBind is an interesting example of Meta’s AI strategy that is focused on open-sourcing the models. This is a quote from Mark Zuckerberg in the Q1 2023 earning call on a question from an analyst that related to open source and AI:

We can see that Meta's approach to open sourcing is set to foster an intriguing dynamic in the AI industry. Mark Zuckerberg's comments underscore a commitment to openness, which starkly contrasts with the strategies of other tech giants like Microsoft, Amazon, and Google. These companies often maintain a proprietary hold over their software infrastructures, primarily because they offer cloud computing services as a key part of their business models.

Meta, on the other hand, isn't in the cloud service business. They see more value in having the industry standardize the tools they use so that they can benefit from improvements made by others and potentially lower costs via wider use. This strategy could lead to more open-source work at the infrastructure level while the product work remains proprietary and unique to Meta's offerings.

By open-sourcing AI models like ImageBind, Meta pushes for an open AI ecosystem, encouraging collective progress in AI research and development. This approach might stimulate a shift in industry practices, possibly influencing other players to adopt a similar open-source strategy to contribute to and leverage collective advancements in AI.

Overall, Meta's strategy could accelerate innovation, democratize access to advanced AI tools, and reshape the competitive landscape in the AI industry. The open-source approach could potentially lead to greater transparency, collaboration, and standardization, thereby creating an environment conducive to rapid advancements in AI technology.

[2] Airbnb Q1 Earnings: Revenue Beat Overshadowed by Booking Miss, Shares Sink 9%

Airbnb (ABNB) reported its Q1 2023 earnings, revealing a slight beat on revenue but a shortfall in bookings. The miss in bookings led to a 9% decline in the company's shares in after-hours trading, despite announcing a new $2.5 billion stock buyback. Airbnb's revenue was $1.8 billion, slightly above the estimated $1.79 billion, but the Nights and Experiences Booked fell short of expectations.

More:

• Actual Q1 revenue: $1.8 billion, versus estimated $1.79 billion

• Actual EPS: 18 cents, versus estimated 12 cents

• Nights and Experiences Booked: $121.1 million, versus estimated $122.4 million

• Q2 Revenue Guidance: $2.35-$2.45 billion, versus estimated $2.42 billion

• Airbnb's shareholder letter highlighted 19% growth in bookings in Q1 2023 compared to the previous year, despite macroeconomic uncertainties

_{Tags: #Airbnb, #Financial report, #Stocks}

Read more about it at —> (read time: 4 minutes)

Insight:

We see this as a market-level impact event. While Airbnb outperformed revenue expectations, the miss on bookings—a key performance indicator—significantly overshadowed this victory. Despite macroeconomic uncertainties, Airbnb grew in active bookers, indicating loyalty from returning guests and a growing base of first-time bookers. The miss on bookings raises questions about Airbnb's ability to capitalize on this loyalty and growth, which could have implications for future quarters.

[3] Meta Shifts Earnings Model: Creators to be Paid Based on Reels' Performance

In a fresh twist, Meta is testing a new payout model for its Ads on Reels monetization program. The proposed model will see Facebook creators compensated based on the performance of their public Reels, rather than ad earnings. The social media giant also plans to introduce the same payouts to Instagram in the forthcoming weeks.

More:

• Meta's new payout model is in the testing phase

• Creators will be compensated based on the performance of their public Reels, marking a shift from the traditional ad earnings model

• Meta plans to extend this payout model to Instagram in the near future

_{Tags: #Big tech, #Meta, #Streaming}

Read more about it at —> (read time: 3 minutes)

Insight:

We believe this move by Meta signals a company-wide impact. The shift in its monetization strategy represents an effort better to align creators' incentives with the platform's goals. By paying creators based on the performance of their content, Meta seems to be encouraging the creation of high-quality, engaging content that will boost user interaction and time spent on its platforms.

This change could further intensify the competition among social media platforms for creators, who are increasingly seen as vital for user engagement and platform growth. If successful, this payout model could prompt other platforms to reconsider their monetization strategies, potentially leading to an industry-wide shift.

[4] Big Tech Mergers Under Fire: Regulators on Both Sides of the Atlantic Tighten Scrutiny

Big Tech acquisitions are facing unprecedented scrutiny as regulators globally flex their antitrust muscles. This comes in the wake of the UK's veto on Microsoft's proposed takeover of Activision. Policymakers and competition officials are eager to examine multibillion-dollar deals more closely, signaling a shift in the regulatory landscape towards tougher enforcement.

More:

• UK regulators recently blocked Microsoft's proposed $69 billion takeover of Activision.

• Policymakers and competition officials globally are eager to scrutinize multibillion-dollar acquisitions more carefully.

• Microsoft, along with other Big Tech companies, face a new reality of stricter antitrust enforcement.

• Regulatory changes in the EU and UK aim to address the concentration of market dominance online.

• Other pending deals, such as Amazon's purchase of iRobot and Adobe's tie-up with Figma, are also under scrutiny.

_{Tags: #Big tech, #Regulation, #M&A, #Microsoft}

Read more about it at —> (read time: 10 minutes)

Insight:

We identify this development as having a significant industry-wide impact. There's a clear signal that the era of laissez-faire towards Big Tech mergers is ending, with stricter antitrust enforcement taking its place. This shift is driven by regulators' concerns about the concentration of market power online and their eagerness to prevent the formation of monopolies that could stifle innovation and competition.

This regulatory tightening could profoundly impact Big Tech's growth and acquisition strategies, forcing these companies to navigate a more complex and uncertain regulatory environment. If these changes become the norm, they could reshape the tech landscape, affecting not just the tech giants but also startups and other companies that could be potential acquisition targets.

⚡️Startup & technology bites

Brief news summaries from the world of startups and tech that provide quick and easy-to-consume information.

Microsoft 365's Copilot, an AI-powered digital assistant based on OpenAI's GPT-4, is expanding preview access with new features and a paid early access program for 600 global customers. The new capabilities include a Semantic Index feature and improved accuracy in reporting internal company data. (link)

Nvidia reportedly set to launch mid-range GPUs, the GeForce RTX 4060 Ti and 4060, this May, adding a 16GB option for RTX 4060 Ti to cater to high VRAM demands of AAA games. (link)

Amazon launches Amazon Anywhere, an immersive shopping platform, allowing in-game purchases of real items, debuting in Niantic's AR pet simulator mobile game, Peridot. (link)

AI company and smart device maker, Mobvoi, is set to hold a Hong Kong IPO, potentially raising $200 million to $300 million. The Beijing-based firm, known for its AI products including smartwatches and smart speakers, last month launched its own AI large language model. (link)

Twilio Inc. anticipates Q2 revenue below analysts' expectations, causing shares to drop. Despite a 15% YoY revenue increase to $1.01 billion in Q1, the firm recently laid off 17% of its workforce to cut costs. They've also initiated a $1 billion share repurchase program. (link)

Palantir's Q1 revenue rose 18% YoY to $525M, surpassing the $506M estimate. The company also announced the launch of its AI Platform for private datasets in May 2023, leading to a 23% jump in shares. (link)

PayPal's Q1 shows a 10% YoY rise in total payment volume to $354.5B, net revenue up 9% to $7.04B, and a 41% increase in operating income to $1B, despite a 5%+ drop in shares. (link)

Affirm's Q3 results show a 7% YoY increase in revenue to $381M, a net loss of $205.7M, and a rise in gross merchandise volume to $4.6B, despite an 8%+ drop in shares. (link)

EA reported Q4 revenue of $1.87B, a 2.7% YoY increase, alongside an 11% rise in net bookings to $1.95B. However, it recorded a $12M net loss compared to a $225M net income YoY. (link)

🔍 Extra

A mix of learning links, software products, deals, and other interesting content.

Venture Capital (VC) firms' own fundraising has hit a ten-year low, indicating a potential tightening of investment in the coming quarters, despite funds remaining from previous record years. (link)

In 2023, corporate VCs have increased their lead in startup funding rounds, including Microsoft's $10 billion investment in OpenAI, as other venture investors scale back. This reflects a shift in investment strategies and time horizons. (link)