Alibaba Utilizes AI for Multimedia Understanding, Tech Equity Inflows Boost S&P, Nasdaq and more

Pivot 5: 5 stories. 5 minutes a day. 5 days a week.

1. Alibaba Pioneers ChatGPT-inspired AI Tech to Advance Multimedia Content Understanding and Translation

Alibaba has started implementing its ChatGPT-like technology, dubbed Tongyi Qianwen, as it ramps up its AI pursuits.

This language model will be integrated into Tingwu, Alibaba's digital assistant, and it is designed to analyze multimedia content and generate text summaries. Now available for public testing, the AI assistant will first be deployed on DingTalk, Alibaba's business-focused messaging service. Further enhancements to Tongyi Tingwu are expected later this year, including real-time English-Chinese translation for multimedia content. This service will be a Chrome web browser plugin. Alibaba's spokesperson confirmed plans to collaborate with corporate cloud clients to create customized AI solutions using the Tongyi Qianwen model.

This rollout takes place amid AI-focused efforts by Chinese tech companies in the face of a sluggish domestic economy and tighter regulations.

Read the full story here

2. Record Inflows into Tech Equity Funds Drive S&P 500, Nasdaq to Nine-Month Highs, BofA Data Reveals

In the week to Wednesday, technology equity funds witnessed a record-setting inflow, BofA Global Research data shows.

The tech stocks experienced an inflow of $8.5 billion, contributing to a general inflow of $14.8 billion for stocks, the largest since February. Nasdaq and S&P 500 reached nine-month closing highs, partly driven by Nvidia's shares' 30% rise over three sessions. BofA analysts, however, remain bearish due to potential liquidity tightening from rising interest rates. They also highlight the market's return to companies with large margins and high price-to-earnings ratios, with seven stocks accounting for 8.8% of the S&P 500's 10% YTD return.

Inflows also increased for cash funds, while gold funds saw outflows. BofA suggests a contrarian trade: buying Hong Kong's Hang Seng Tech Index and selling the Nasdaq 100, expecting a surprise stimulus from China.

Read the full story here

3. AI Tools Gain Momentum in Identifying Acquisition Targets and Startups for Investment

Venture capital funds, private equity groups, and accountancy firms have started employing advanced artificial intelligence (AI) to identify potential acquisition targets and promising startups.

Notable entities using this tech include KPMG, Coatue, and venture capital firm Headline. They utilize generative AI to analyze a company's growth potential. KPMG uses the tech behind OpenAI's ChatGPT to create a system that assists its staff, while Coatue's software integrates generative AI to sift through sellside research and other data. PitchBook's AI-powered "VC exit predictor" boasts a 75% accuracy rate, while firms like Headline and Moonfire Ventures use AI to assess investment targets based on various growth metrics.

Despite its increasing use, the impact of AI on traditional roles in the sector remains a topic of debate, with some arguing it serves as a co-pilot rather than a replacement for human judgment.

Read the full story here

4. MIT's CSAIL Develops Scalable, Self-Learning Language Model Challenging Dominance of Large Models

Researchers at the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) have developed a scalable, self-learning language model that challenges the notion that larger models offer superior capabilities.

The model, called SimPLE (Simple Pseudo-Label Editing), uses self-training, allowing it to learn from its own predictions and eliminating the need for additional annotated training data. This model significantly improves performance across a range of tasks, outperforming notable models such as Google’s LaMDA, FLAN and other GPT models. The team argues that their research challenges the idea that larger models are inherently superior, and highlights the potential of smaller models as equally powerful and environmentally sustainable alternatives. The SimPLE model focuses on textual entailment, a strategy that significantly enhances the model's versatility and adaptability.

This approach offers a cost-effective, scalable, and privacy-preserving alternative in the AI landscape.

Read the full story here

5. AI Poised to Tackle $5 Billion Synthetic Identity Fraud Challenge

Synthetic identity fraud, projected to cost global financial and commerce systems $5 billion by 2024, saw losses amount to 5.3% of all digital fraud in 2022, a 132% increase. Sontiq, a TransUnion company, linked data breaches with synthetic identity fraud, finding outstanding balances attributed to synthetic identities in the U.S. rose to a record $4.6 billion in 2022.

Fraud destroys customers' trust, further aggravated by the fact that 10% of credit and debit card users have experienced fraud in a year. Synthetic identity fraud is challenging to detect as it involves using comprehensive personal data to create fake identities. Fraud detection models miss 85-95% of likely synthetic identities, requiring enhanced tools and apps.

AI plays a crucial role in combating synthetic identity fraud, and five ways it assists include designing machine learning (ML) into the core code base, reducing latency of identifying synthetic fraud via cloud services, integrating user authentication with identity proofing, using ML-based risk scores to reduce false positives, and applying predictive analytics for real-time anomaly detection.

Read the full story here