Last week in AI: AI wearables struggle, Huawei’s new AI chip, Elon Musk’s Grok 2 launch, Google rolls out Pixel 9

AI wearables struggle: Humane AI Pin faces high returns and poor sales

The Humane AI Pin, released in April, faces significant challenges, with more customers returning the product than buying it. The company initially aimed to sell 100,000 units but has reportedly only shipped 10,000. To add insult to injury, 3,000 units—equivalent to about $2.1 million in sales—have been returned, leaving only about 7,000 Humane AI Pins in customers’ hands, a staggering 93% below the company’s original goal.

This outcome shouldn’t come as a surprise. Shortly after its release, negative reviews started flooding in, with prominent tech reviewer Marques Brownlee calling it the worst product he’s ever reviewed.

This trend reflects what we’ve seen across the AI wearable space; these products receive much more negative feedback and reviews than positive feedback and reviews. This is the consequence of the approach many companies are taking to building in the space.

Most are trying to create AI wearables marketed as “replacements” for items like smartphones. Unfortunately, these replacements often introduce more friction into workflows rather than reducing it. As a result, companies are launching products that feel unfinished, which really only attracts technologists willing to demo the product but doesn’t make them or consumers interested enough to continue using or actually going out to buy the product.

Huawei challenges Nvidia with new AI chip

Huawei Technologies has announced its latest chip, the Ascend 910C, which the company claims is comparable to Nvidia’s (NASDAQ: NVDA) H100—one of the highest-performing chips on the market, especially for AI workloads.

This move is likely inspired or catalyzed by U.S. sanctions that restrict Nvidia from selling its most advanced chips in China due to national security concerns. The AI race goes far beyond companies creating AI models; the hardware these companies need to train and run the models may be even more important than the models themselves. While Nvidia has dominated this space with 88% market share, Huawei may be a formidable competitor if its AI Chip is really up to par with Nvidia’s.

Elon Musk’s xAI launches Grok 2

xAI, the artificial intelligence arm of Elon Musk’s X.com, recently announced the launch of the Grok 2 beta. Grok is an AI chatbot native to x.com (formerly known as Twitter). This new version of Grok includes an image generator, and the company claims it outperforms other models on the market, such as Claude 3.5 Sonnet and GPT-4 Turbo. Grok 2 beta is currently available to X Premium and Premium+ users.

One of Grok’s unique selling points is its less restricted, less censored outputs compared to other chatbots, often resulting in humorous, witty, or even crass responses. However, I believe we’re reaching a point in the market where we need to ask ourselves: how many chatbots do we really need? Most AI chatbots aren’t functionaly better or substantially different from one another, besides OpenAI’s ChatGPT-4o, which dominates the space and controls the lion’s share of the market.

Remember that training and running large language models (LLMs) is incredibly expensive, and not even the largest player—OpenAI—has turned a profit yet. If I had to guess, the AI chatbot development trend from tech companies is unsustainable. We’re likely to see companies either consolidate or sunset their AI chatbots as the costs of keeping them online continue to rise while profits remain elusive.

Google bets on AI with Pixel 9

At Google’s (NASDAQ: GOOGL) Pixel 9 launch event, the company focused more on AI than its latest smartphone. From the start of Google’s senior Vice President of Platforms & Devices Rick Osterloh‘s presentation until about the 30-minute mark, he didn’t even mention the Pixel 9, which was supposed to be the centerpiece of the show. Instead, Osterloh highlighted artificial intelligence, particularly Google’s Gemini, and how it would be integrated into the Pixel 9’s software and hardware.

“A few months ago at Google I/O, we shared a broad range of breakthroughs to make AI more helpful for everyone. We’re obsessed with the idea that AI can make life easier and more productive for people. It can help us learn. It can help us express ourselves. And it can help us be more creative. The most important place to get this right is in the devices we carry with us every day. So we’re going to share Google’s progress in bringing cutting-edge AI to mobile in a way that benefits the entire Android ecosystem,” said Osterloh in his opening remarks before discussing features like Gemini Live, an AI tool for brainstorming or interview practice, along with Gemini’s image editing, text analysis, and upcoming computer vision capabilities.

Google may have pushed AI so hard at this event due to recent earnings calls, during which analysts questioned when Google’s significant investments in AI would begin to turn a profit. Maybe they made AI a focal point of the Pixel 9 to further try to justify Google’s massive spending on artificial intelligence.

However, will this “AI marketing” strategy have its desired effect of driving sales and getting people excited to use AI features regularly? We still haven’t seen proof that consumers demand AI products outside of chatbots. Even long-standing AI products like Apple’s (NASDAQ: AAPL) Siri have struggled to gain mass adoption—which is probably a bad sign for Gemini Live.

In order for artificial intelligence (AI) to work right within the law and thrive in the face of growing challenges, it needs to integrate an enterprise blockchain system that ensures data input quality and ownership—allowing it to keep data safe while also guaranteeing the immutability of data. Check out CoinGeek’s coverage on this emerging tech to learn more why Enterprise blockchain will be the backbone of AI.

Watch: Understanding the dynamics of blockchain & AI