AI Megatrend Re-Accelerates as Analysts Highlight Stock Picks

AI systems can quickly process large amounts of text, but a new study suggests they struggle when financial inputs are embedded in charts and visuals rather than provided as clean, labeled data. The findings come as investors weigh how to use AI tools alongside human analysis.

Study finds accuracy drops when numbers are embedded in charts

The study, from Mercor, an AI-powered hiring platform that uses large language models to screen resumes, conduct interviews, and match talent with companies, tested several advanced AI models on real financial tasks. The models included Claude Opus 4.6, GPT-5.4, and Gemini 3.1 Pro.

According to the researchers, when the data was clean and written out in text, the models answered correctly roughly 75% of the time. However, when the same numbers were embedded in charts and visuals—similar to those used in everyday reports—accuracy dropped sharply.

The study attributes the decline to errors in reading chart inputs, including misinterpreting the chart axis or selecting the wrong data point. Once an initial mistake is made, subsequent outputs were also incorrect.

Human analysts retain an advantage in real-world chart reading

The article argues that investing is not typically done with perfectly labeled spreadsheets. In that environment, human analysts may have an advantage in interpreting visual information and avoiding errors that can cascade from a single misread input.

Market focus shifts to AI infrastructure and earnings signals

Separately, tech investing expert Luke Lango said he acted after a ceasefire was announced, describing it as the resumption of an “AI bull market.” He recommended five stocks across two themes: the AI Infrastructure Buildout and the Hard Assets Boom Cycle.

One highlighted pick was KLA Corp. (KLAC). The article describes KLA as providing inspection and metrology tools used to ensure semiconductor chips function correctly, noting that even small defects can have major financial impact. It also states that the stock traded sideways during “Operation Epic Fury” before resuming its climb, while the S&P 500 remained in negative territory over the same period.

Inflation and earnings expectations: PCE and CPI data

On Thursday, the Bureau of Economic Analysis released the February Personal Consumption Expenditures (PCE) index, described as the Federal Reserve’s preferred inflation gauge. Core PCE was 3.0% year-over-year, matching consensus estimates. Headline PCE was 2.8% annually. On a monthly basis, both core and headline rose 0.4%.

On Friday, the government released Consumer Price Index (CPI) data. Consumer prices rose 0.9% in March, driven by nearly an 11% jump in energy costs. Overall prices were up 3.3% from a year earlier. “Core” prices excluding food and energy rose 0.2% from the prior month and 2.6% from a year earlier.

The annual inflation rate rose to 3.3%, described as the highest level not seen in nearly two years. The article also notes that Q4 GDP growth was revised down to 0.5% annualized from an initial 1.4% estimate.

Expert view: stay focused on fundamentals

Investing legend Louis Navellier said he was not satisfied with the inflation results but emphasized a focus on stock fundamentals. In his commentary to Growth Investor subscribers, he pointed to “wave after wave of positive analyst earnings revisions” and said relative strength and institutional buying pressure continued from quarter-end into the current week.

Navellier also argued that investors should not move to the sidelines, describing a strategy of maintaining exposure to fundamentally strong stocks. He cited Growth Investor stocks with 37.3% average forecasted annual sales growth and 144% average forecasted annual earnings growth.

AI infrastructure stock example: Ciena

The article says Navellier is also steering subscribers toward AI infrastructure stocks. It highlights Ciena (CIEN), described as a connectivity provider of optical networking systems and software supporting adaptive networks and increasing bandwidth demand.

Since the recommendation mentioned in late February, the article states CIEN is up almost 40%, while the S&P 500 has not broken even over the same period. It also reports Ciena’s revenue expectations: total revenue of about $1.5 billion versus $1.07 billion in the same quarter a year ago. For fiscal year 2026, the company anticipates total revenue between $5.9 billion and $6.3 billion, representing 23.7% to 32.1% annual revenue growth.

Bottom line

While AI may excel at processing text, the Mercor study suggests it can misread chart-based inputs, reducing reliability when numbers are presented visually. Meanwhile, the article frames current investor attention around inflation data, revised growth expectations, and earnings revisions, alongside continued interest in AI infrastructure stocks.