Urgent.News

What's breaking now, across thousands of outlets.

Tech

SoftBank swings to $7 billion net loss in second quarter as tech selloff weighs on results

SoftBank experienced a significant setback in the second quarter, reporting a net loss of $7 billion, compared to a profit of $2.5 billion in the same period last year. The loss was primarily attributed to the firm's substantial investments in artificial intelligence, particularly its $64.6 billion stake in OpenAI. Despite forecasts from three analysts ranging from a profit of $233 billion to a loss of $113 billion, SoftBank's earnings remain highly unpredictable.

The company's heavy involvement in AI, driven by its significant investments, has left it vulnerable to the recent sector-wide challenges. The Democratic Party's victory in the US midterm elections and growing regulatory concerns have further intensified investor anxieties regarding the potential impact of AI on returns. Additionally, SoftBank's shares have plummeted, losing nearly half their value since mid-June.

The recent FBI investigation into Nvidia CEO Jensen Huang, which surfaced amid allegations of espionage by a Chinese national, has added another layer of uncertainty to the already tumultuous AI landscape. It is worth noting that SoftBank was once a major investor in the AI chipmaker but currently holds no stake in the company.

Written by urgent.news from Investing.com's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

This story

This is one outlet's version. Read the fullest account.

Read the original at cnbc.com →

More in Tech

Every answer shows its source

An agent that answers questions is only as useful as people's confidence in its answers. The most reliable way to earn that confidence is to make every answer traceable. Cite, then show.

  • Every answer must link to an approved document's specific passage.
  • Update policy documents before using them in responses.
  • Decline answers for unaddressed questions, redirect to appropriate team.

Serving a heavy model without blowing up the server bill

An AI model is heavy and demanding. If served without precaution, it blows up response times and server costs. On a SaaS audio analysis that I designed and put online alone, it was necessary to keep…

  • Le modèle IA est hébergé sur un serveur propre pour contrôler les coûts.
  • Le modèle est chargé une seule fois au démarrage via loadmodel() en FastAPI.
  • Les calculs sont effectués de manière asynchrone pour éviter les blocages du serveur.

More from Tuesday 10 November →