On Wednesday, Humane announced a $200 price cut for its flagship product, AI Pin. The Bay Area startup, founded by two former Apple executives, was reportedly struggling to sell the product, which was launched in April for $700.

In a message sent to its email list, Humane is announcing its recent returns policy. “Ai Pin starts at $499 and includes the first month of Humane Plan,” the corporate writes. “With a 90-day return period, it’s completely risk-free to try.

The device was poorly received by reviewers. According to reports, AI Pin returns began in August get ahead of salesleaving roughly 7,000-8,000 devices within the hands of users.

The company reportedly began considering a sale in May, admitting to poor reviews and poorer sales.

Andreessen Horowitz has a large cluster of Nvidia H100 GPUs which can be helping a portfolio of artificial intelligence startups meet their computing needs, the enterprise capital firm confirmed for the primary time on Wednesday. The program, called “Oxygen,” allows portfolio firms to coach or operate AI models without negotiating market rates.

A16Z’s Oxygen cluster gives startups a respiration space, so to talk, to compete with larger tech firms – like Google, Meta and Microsoft – in constructing large AI models. For several years, these firms have been engaged in a bidding war for AI industry standard GPUs, especially the H100, often winning contracts by promising larger and longer contracts. This leaves many AI startups out within the cold as they might not have the opportunity to tackle the massive contracts needed to secure GPUs, unlike the OpenAI and Anthropics of the world.

“It started with the realization that many of the AI ​​developers we serve every day had a common problem: We found ourselves in the middle of a supply crisis when the processing power of the Nvidia H100 card was in short supply,” said A16Z partner Anjney Midha, who helped create this system Oxygen in a podcast transcript provided to TechCrunch. “As startups, large clouds de-prioritized them in favor of larger customers, which was really difficult for them.”

Infrastructure can be a superb way for A16Z to draw latest startups. Companies constructing AI models often need unlimited access to very large compute clusters for brief training runs, but this isn’t any longer needed once the models are trained. Inferring AI models typically requires less computation than training, unless the usage of the AI ​​model explodes. A16Z’s Oxygen cluster gives startups the flexibleness to have access to GPUs once they need them and not using a long-term commitment to a cloud provider or an enormous outlay of cash. Instead, they provide A16Z a stake in their business in exchange for (amongst other things) low GPU rental prices.

But A16Z aren’t the one investors who promise startup founders GPUs along with financial capital and guidance in exchange for shares in their firms. Investment partners Nat Friedman and Daniel Gross offer their startups access to a 4000 GPU cluster in accordance with Forbes, it was named the Andromeda Cluster. Y Combinator also offers startups a GPU cluster for training purposes through partnerships with various cloud service providers, most recently Google Cloud.

A spokesman for Andreessen Horowitz declined to comment on the dimensions of the Oxygen cluster, nonetheless A16Z could have the most important GPU war chest of any enterprise capital firm. Information in July it said Oxygen could contain greater than 20,000 GPUs.

At one point within the podcast, Midha claims that the Oxygen cluster could ease the pressure on AI startups to boost funds at inflated valuations to pay their computing bills. This is one other way oxygen advantages A16Z. Instead of investing more in a startup so it may buy GPUs from, say, Microsoft, A16Z can offer just the GPUs and invest at a cheaper price.

A16Z sees oxygen as a core value proposition for its startups, which can proceed to thrive so long as the AI ​​boom continues as predicted. This is probably going why the startup made the large investments required to secure these chips.

While Amazon Go pioneered the completely touchless store concept, others like 7-Eleven and Walmart have entered this territory. The area is growing, in part because several startups – similar to Standard Cognition, Zippin, AiFi, Grabango, Trigo – have expanded to supply their technologies to retailers.

But in Europe, a Portuguese startup Sensei is developing a contactless store platform and has now secured a €15 million Series A funding round led by BlueCrow Capital. New investors Lince Capital, Explorer Investments and Kamay Ventures (the investment arm of Coca-Cola and Arcor Group), in addition to existing investors Metro AG and Techstars Ventures, also participated in the meeting.

In 2021, based in Lisbon Sensei raised a seed round value $6.5 million (€5.4 million). Then he managed the financing Just like these ventures AND The capital of Iberiswith participation from 200M fund.

The startup currently plans to create 1,000 fully autonomous sales points by 2026. It already serves clients in Portugal, Spain, France, Italy and Brazil, and plans to expand into Central and Northern Europe.

Using computer vision, AI-powered sensors and real-time algorithms, the Sensei system mechanically updates the shopper’s basket and displays a listing of things ready for payment, with identity and privacy protected.

Touchless stores reduce checkout costs, prevent out-of-stocks and provide retailers with real-time visibility into store operations.

During the decision, Vasco Portugal, CEO and co-founder of Sensei, said: “We are growing, especially in the last year, we have almost doubled the number of stores. So we opened stores in a short time. We are now present in five locations.”

“There are two problems in the retail industry. The customer experience sucks. Second, it is extremely difficult to process all sales information in real time. It’s really about shop automation, like automotive automation and factory automation. I believe it is a natural change,” he added.

António de Mello Campello, partner at BlueCrow Capital, added in an announcement: “Artificial intelligence… is transforming several industries and we believe that retail represents one of the biggest opportunities and will soon be disrupted. Sensei has proven to be one of the best in the industry.”

Taking all this into consideration, competition in space is big.

Standard Cognition raised $239.4 million, Wheat $199 million, Grabango $93.8 million, AiFi $87.1 million and Zippin $44 million.