Over the past week, the world has learned a lot about the type of people who invest in Apple (AAPL) .
The tech leader has been in full focus since its shareholders voted against scrapping its diversity, equity and inclusion (DEI) policy, refusing to bow to pressure from President Donald Trump to join big tech’s anti-DEI push.
💵💰Don’t miss the move: Subscribe to TheStreet’s free daily newsletter 💰💵
This comes just after Mr. Trump praised the company for announcing plans to invest 500 billion dollars to create 20,000 U.S. jobs in the coming years. The president is likely feeling conflicted, despite the fact that Apple CEO Tim Cook, a former Democrat, donated $1 million to his inaugural fund.
Cook’s decision to make the donation sparked ire among Apple consumers, who expressed displeasure that he would try to curry favor with Trump. Although interest quickly shifted away from this news moment, a recent incident may have inadvertently revealed a key detail regarding Apple’s consumers.
Edward Berthelot/Getty
Apple is on the spotlight after an embarrassing AI error
Over the past few months, Apple has kept its followers watching closely for key updates on products, including Apple Intelligence. The artificial intelligence (AI) platform has been in full focus but not always for the right reasons.
In December 2024, Apple sent a news notification summary to iPhone users, falsely claiming that Luigi Mangione, the man accused of shooting and killing UnitedHealthcare (UNH) , had committed suicide. Now it seems to have made an error that is just as embarrassing, if not more so.
Related: Apple AI tool falsely announces death of trending murder suspect
According to reports, Apple’s voice-to-text function briefly substitutes the word “Trump” for “racist” as well as other words that begin with an “r” consonant. The company attributes this vocabulary mishap to its “speech recognition models displaying words that have some level of “phonetic overlap.”
Conservative personalities, such as Alex Jones, have accused Apple of “subliminal programming,” alleging that the company has engaged in underhanded activity as a means of harming Trump.
However, actual AI experts insist that the incident is due to a technical malfunction, as Apple stated, and is simply another example of the technology simply making a mistake. The Financial Times reports that it tested this Apple feature and found the Trump suggestion appeared on words besides “racist.”
However, this error does reveal something important.
Brian Krassenstein, a popular political and tech commentator, shared a video of the voice command on his X account, noting that Apple’s model is reflecting how many people speak and associate words.
- Netizens find creative ways to troll DOGE on behalf of federal workers
- Controversial author publishes shocking AI conversation with ChatGPT
- Elon Musk’s use of AI could spell doom for a powerful industry
“Voice recognition systems learn from massive datasets, predicting words based on common associations,” he states. “Since “Trump” and “racist” frequently appear in the same context, the system momentarily links them.”
Apple’s software glitch shows that the industry is still evolving
Trump has not responded to news of this Apple AI situation, but the company has promised to fix it, stating “We are aware of an issue with the speech recognition model that powers dictation and we are rolling out a fix today.”
Related: Apple makes a big announcement, but will it deliver
Scott Stephenson, Founder and CEO of Deepgram, provided context on Apple’s Trump-related glitch, stating:
“AI is only as smart as the data it’s trained on. Voice recognition should be about understanding, not assuming. This is a reminder that companies need to constantly refine their models to avoid bizarre and potentially harmful mix-ups.”
He adds that when it comes to advanced AI tools like Apple Intelligence, intended to benefit consumers, the primary goal should be to establish trust in this fast-growing technology.
Related: Veteran fund manager unveils eye-popping S&P 500 forecast
