AI agents are advancing in sophistication. They are transitioning from merely answering questions to independently executing intricate, multi-step tasks.

However, before these agents can be relied upon to arrange travel or perform financial analyses for users, model developers and the startups creating such agents seek to verify their reliability across a broad spectrum of situations.

AI laboratories frequently utilize benchmarks to demonstrate their model’s capabilities, but a top score, even on an agent-focused benchmark, does not genuinely confirm that an AI can accurately carry out diverse complex real-world tasks.

Patronus AI, a startup established in 2023 by former Meta AI researchers Anand Kannappan and Rebecca Qian, aids model creators and companies in refining models for this purpose by constructing simulated digital environments to assess the agents’ performance.

The San Francisco-based startup appears to be addressing a crucial issue. Nearly every leading AI lab and many emerging startups are now clients, according to Glenn Solomon, a managing director at Notable Capital, who notes that demand for the company’s simulated environments is almost boundless.

Patronus has seen its revenue expand 15 times over the last year, attracting considerable investor interest. On Thursday, the firm declared a $50 million Series B funding round led by Greenfield Partners, with contributions from Notable Capital, Lightspeed, Datadog, and Samsung. This funding round elevates the company’s total financing to $70 million.

Patronus employs what it terms “digital world models” to replicate websites and internal systems. Within these environments, agents are rigorously tested post-training through reinforcement learning, which progressively rewards successful task execution and penalizes mistakes.

AI laboratories recognize significant value in these digital simulations as they provide agents with opportunities to engage in varied, occasionally unpredictable, scenarios. The firm likens its methodology to Waymo’s strategy for training autonomous vehicles by first creating synthetic environments to evaluate cars against rare risks, like severe weather or a child chasing a ball.

The distinction with AI agents lies in their tendency to take shortcuts, resulting in incomplete task execution. “Patronus excels at identifying these shortcuts and ensuring that the models are held accountable,” Solomon stated.

Currently, Patronus is offering its simulated digital environments for software development and finance, but according to Kannappan, these are only the initial areas of focus.

“Currently, we are primarily concentrated on problems that can be verified, specifically those that can be immediately checked and confirmed, but numerous other domains exist that are either highly non-verifiable or extremely challenging to confirm,” he remarked.

Just because these operations are verifiable does not imply they are straightforward. “We aim to create environments in which agents can operate continuously for 10 hours, 10 days, or even 10 weeks,” Kannappan noted.

Regarding competitors, Patronus sees itself as mainly contending against the internal teams AI labs have already assembled to assess agent performance. While human-data companies like Mercor and Surge assist model developers with reinforcement learning, Patronus differentiates itself by evaluating agent behavior without human intervention.

Only a short time after Apple declared price hikes across its hardware offerings, Microsoft announced that its Xbox game consoles will also see price increases. Alongside this, the company revealed the discontinuation of its 2TB model.

Effective August 1, prices for Xbox consoles will rise globally. The cost of the 512GB models will go up by $100, while the 1TB variants will increase by $150. 

The pricing adjustments are detailed as follows: 

• Xbox Series S 512GB is rising from $399 to $499.
• Xbox Series S 1TB is increasing from $449 to $599.
• Xbox Series X 1TB Digital is going up from $599 to $750.
• Xbox Series X 1TB Disc is climbing from $649 to $800.

According to the company, these increases are propelled by escalating costs of memory and console storage, which are currently more than 2.5 times higher than previous figures. Microsoft cautioned that these prices might potentially double by fall 2027. This change comes less than a year after the company raised Xbox prices in the U.S. last October.

This announcement follows Apple’s own series of price increases affecting products like Macs and iPads. Apple referenced the same industry-wide pressures, indicating the surging costs of memory and storage driven by unmatched demand for AI infrastructure and data centers.

Taken together, these sequential announcements highlight the significant impact of the AI surge on the pricing of daily electronics. As tech companies heavily invest in expansive AI systems, the demand for advanced memory and storage chips has surged, constraining supply chains and elevating costs across the industry.

Microsoft sought to cushion the impact by emphasizing financing options and plans to broaden access to more affordable hardware. In its announcement, the company stated it is “developing new programs to offer pre-owned consoles at reduced prices.”

Customers acquiring qualifying Xbox hardware through Microsoft Stores will gain improved access to buy now, pay later programs, while Amazon buyers can be eligible for up to 12 months of 0% APR financing on qualifying purchases.

Moreover, Microsoft now aligns with Sony in asking gamers to spend more, as the PS5 digital version has markedly risen from its launch price of $499 to $599. On the other hand, Nintendo’s price increase for the Switch 2 has been relatively modest, but the competitor may experience pressure to escalate prices further in the future.