Serve is betting that food delivery and access to public markets are the keys to scaling robotics
Backed by Nvidia and Uber, Serve recently raised $80 million to extend its runway through 2026. The company aims to scale from 100 sidewalk delivery robots in Los Angeles to 2,000 bots operating across U.S. cities by the end of this year and hit operational profitability once that fleet is fully deployed. It's a bold play in a space where hardware, logistics, and data all collide.
Today on Equity, Rebecca Bellan caught up with Kashani to unpack how Serve is navigating public markets, scaling real-world robotics by using food delivery as a test ground, and building what it hopes is the future of last-mile delivery.
Listen to the full episode to hear more about:
How Serve went from a Postmates spinout in 2021 to a publicly traded company via reverse merger in 2024.
What it takes to scale a delivery fleet across cities like L.A., Miami, and Dallas, and why Serve isn't launching on college campuses like its rivals.
Why Kashani says Serve's sidewalk bots collect four times more visual data per day than GPT-4's vision model.
How ground robots and drones might work together to finally crack last-mile logistics.
Equity will be back Friday with our weekly news round-up, and special Google I/O coverage from Max. Don't miss it!
Equity is TechCrunch's flagship podcast, produced by Theresa Loconsolo, and posts every Wednesday and Friday.
Subscribe to us on Apple Podcasts, Overcast, Spotify and all the casts. You also can follow Equity on X and Threads, at @EquityPod. For the full episode transcript, for those who prefer reading over listening, check out our full archive of episodes here.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
CNBC
3 minutes ago
- CNBC
Chinese AI firms form alliances to build domestic ecosystem amid U.S. curbs
China's artificial intelligence companies have announced two new industry alliances, aiming to develop a domestic ecosystem to reduce dependence on foreign tech as they seek to cope with U.S. export restrictions on advanced Nvidia chipsets. The announcements were timed to coincide with the three-day World Artificial Intelligence Conference in Shanghai ending on Monday. The conference also showcased a slew of new products, such as an AI computing system from Huawei that experts believe rivals Nvidia's most advanced offering, as well as consumer-friendly products such as several kinds of digital AI glasses. The "Model-Chip Ecosystem Innovation Alliance" brings together Chinese developers of large language models (LLMs) and AI chip manufacturers. "This is an innovative ecosystem that connects the complete technology chain from chips to models to infrastructure," said Zhao Lidong, CEO of Enflame, one of the participating chipmakers. Other manufacturers of graphics processing units (GPUs) in the alliance include Huawei, Biren, and Moore Threads, which have been hit by U.S. sanctions that block them from purchasing advanced tech made with U.S. know-how. The alliance was announced by StepFun, an LLM developer. A second alliance, the Shanghai General Chamber of Commerce AI Committee, aims to "promote the deep integration of AI technology and industrial transformation." Participants include SenseTime, also sanctioned by the U.S. and which has pivoted from facial recognition technology to LLMs. Others are StepFun and another LLM developer, MiniMax, as well as chipmakers Metax and Iluvatar CoreX. One of the most talked about products at the conference was Huawei's CloudMatrix 384 which incorporates 384 of its latest 910C chips and outperforms Nvidia's GB200 NVL72 on some metrics, according to U.S. research firm SemiAnalysis. Huawei's system design capabilities have meant that it has been able to use more chips and system-level innovations to compensate for weaker individual chip performance, SemiAnalysis said. At least six other Chinese computing firms showcased similar "clustering" chip technology. Metax demonstrated an AI supernode featuring 128 C550 chips designed to support large-scale liquid-cooled data centre requirements. Other events included Tencent's unveiling of its open-source Hunyuan3D World Model 1.0, which the company said enables users to generate interactive 3D environments through text or image prompts. Baidu announced what it said was next-generation "digital human" technology that helps businesses to create virtual livestreamers. It features "cloning technology" that can replicate a human's voice, tone, and body language from just 10 minutes of sample footage. Alibaba was among those announcing AI glasses. Its Quark AI Glasses are powered by its Qwen AI model and are due to be released in China by the end of 2025. They will allow users to access the tech giant's map service for easy navigating and to use Alipay by scanning QR codes with voice commands.
Forbes
33 minutes ago
- Forbes
Following Up On President Trump's Idea Of Renaming AI
President Trump included a seemingly ad hoc remark during his AI Summit speech about whether the AI ... More phrase might need an upgrade. In today's column, I explore the ongoing saga of how the moniker 'artificial intelligence' has been received over its approximately seventy-year history, including a recent offhand remark made by President Trump last week during his AI summit speech, suggesting that perhaps it is time to rename AI to something more fitting. Let's talk about it. This analysis of AI breakthroughs is part of my ongoing Forbes column coverage on the latest in AI, including identifying and explaining various impactful AI complexities (see the link here). The Extemporaneous Remark There were several important and impactful announcements about AI made by the White House and President Trump on July 23, 2025, at the Winning the AI Race summit held in the Andrew W. Mellon Auditorium in Washington, DC. In addition to signing three Executive Orders regarding AI, President Donald Trump gave a speech to highlight various present and future aspects of AI. During the speech, he seemed to make an ad hoc remark about the phrasing of AI, which caught the eye of some but was generally overlooked by many as an ad hoc mid-thought commentary. Here's the pertinent portion from his speech (excerpted): Observe that the suggestion involved the consideration that 'artificial intelligence' as a moniker is potentially an understatement of what the technology bodes for. The emphasis seems to be that the word 'artificial' is insufficient to adequately capture the immensity of the technology. The word 'artificial' certainly has varying connotations, both favorable and unfavorable. Sometimes, artificial implies that something is a mere imitation. For example, we might say that a bouquet of plastic flowers isn't real; they are merely artificial. That's one way of interpreting the word. Another even less favorable interpretation is that artificial refers to being pseudo or half-baked, such as saying that a person's smile was artificial. Unpacking The AI Naming You might be wondering how we landed on the AI moniker to begin with. The now-popular phrase 'artificial intelligence' has gone through many ups and downs during its roughly 70-year-long history, but managed to survive and endure. There have even been times when the phrase was highly heralded and placed nearly on a pedestal. Other times have not necessarily been so laudatory. We've seen periods where attempts have been made to either maneuver the phrase into the background or perhaps replace it entirely. How did the phrase get started? By and large, the most agreed-upon official starting point occurred in 1955. It was documented in a famous research proposal entitled 'A Proposal for the Dartmouth Summer Research Project on Artificial Intelligence' by J. McCarthy, M. Minsky, N. Rochester, and C. Shannon, August 31, 1955 (excerpts): The essence of the phrase is that we want to devise computers that can exhibit intelligent behavior equal to or surpassing that of human intelligence. Note that the computer doesn't necessarily have to be some kind of recreation of the human brain. If we could bring forth intelligent behavior by simply using a vast assortment of Legos, that would be fine and provide the intended outcome. Tech Names Change Over Time It is not unusual that the names of technology might change over time. A new technology can start with one name, then others are floated as a potential replacement, and a Darwinian battle ultimately determines which one will get the final nod. The skirmish entails a semblance of what society and culture deem worthy and fashionable. Consider a vivid example that illustrates this principle. Today, we readily and without hesitation refer to the online world as the Internet. There was a time in the 1970s, 1980s, and somewhat into the 1990s when other verbiage was fighting for prominence. You might be vaguely aware that a big initial push was to refer to the online approach as the information superhighway. Lots of variations came and went. There was the info superhighway, data superhighway, and electronic superhighway. Notice that the 'superhighway' remained consistent, while the word preceding it was the item in question. This might be likened to keeping the word 'intelligence' in the AI phrase and seeking to find some other first word besides 'artificial'. Catchier ways of expressing the information superhighway also appeared. There was the infobahn, a clever recasting of high-speed autobahns for cars. Another competing phrase was the World Wide Web, abbreviated as WWW. In the end, we have come to use the Internet moniker and seem to have little memory of the lengthy trials and tribulations of coalescing on this vaunted final phrasing. Is the naming of technology important? Yes, it is, indeed. Per a point made famous by American entrepreneur Jim Rohn: 'Vocabulary enables us to interpret and to express. If you have a limited vocabulary, you will also have a limited vision and a limited future.' Various Names For AI AI has had a topsy-turvy history. Moments have arisen that a belief in striking gold and finally achieving full-on intelligence via computers was considered immediately on the horizon. When those hoped-for accomplishments did not arise, the AI moniker shifted to being more so a sour name rather than a badge of honor. Likewise, AI-adjacent technologies, or some would say subsumed technologies, often rose to prominence and seemed to supersede the AI moniker. During the days of expert systems, some eschewed the AI name in favor of referring to expert systems, rules-based systems, knowledge-based systems, and so on. AI was relegated to the backseat as a naming convention. If anything, AI at that time was used mainly to mention robotics and other physically aimed constructs. That line of technology gradually waned and no longer carried the cachet it initially had. Today, seasoned technologists cheekily refer to that tech as GOFAI, good old-fashioned AI. The irony is that there is a chance that the technology will make a comeback. For more on a potential resurgence of rules-based approaches that combine with contemporary AI, known as hybrid or neuro-symbolic technology, see my coverage at the link here. Ideas On Replacing The AI Name Let's suppose that we wanted to find a moniker that would replace the 'artificial intelligence' phrase on a go-forward basis. There are some crucial factors to consider. The new phrase needs to be relatively short. Perhaps two words in length, maybe three at the most. A one-word phrase probably wouldn't be sufficient. The use of four words would be clunky and unwieldy. The moniker would seem to need to include the word 'intelligence' or at least some similarly invocative word. Thus, we would have one word in front of intelligence, or perhaps one word after intelligence, or maybe two such words. It must be catchy and roll off the tongue. The new phrase must be easy to say and easy to write. The nature of its abbreviation would also be an important element in overall acceptance. We want to avoid using a phrase that is already in common use. Any pre-existing phrase will already have baggage associated with it. The use of an existing phrase for an entirely new connotation would undoubtedly sow confusion. There might also be conflicts with trademarks or other awkward entanglements. Floating Some Possibilities Another facet would be to reject the use of the word 'artificial' due to its variation in meaning. Assume that we are desirous of getting rid of artificial from the moniker. We aim to formulate a two-word or three-word catchy phrase that could live up to the assumed bright future of the technology. Noodle on that puzzle. One possibility is to use Machine Learning (ML) or possibly Machine Intelligence (MI) as a replacement moniker. Great, maybe we've found a handy-dandy replacement. Easy-peasy. Sorry to say that it turns out that both of those have already been tried. Generally, in the in-between status of expert systems and modern-era generative AI, there was a great deal of attention on machine learning and devising machine intelligence. The ML moniker especially gained a lot of traction. Nowadays, ML and, to some degree, MI are considered backroom words that are mainly used amongst techies. The customary viewpoint now is that ML and MI are subsets of the broader field of AI. Not everyone concurs with that point. Some believe that it is insulting or perhaps demeaning to classify those vital areas in that manner. Keep The Juices Flowing Let's do more brainstorming. We could try using the phrase Computational Intelligence (CI). It retains the desired word 'intelligence,' and it is a succinct two-word phrase. The issue is that the word 'computational' seems exceedingly flat. You aren't likely to be stirred by the word. The phrase lacks sizzle as a replacement for the AI phrase. What about Synthetic Intelligence (SI) or possible Algorithmic Intelligence (AI)? People generally like the word 'synthetic' as used in the phrases of synthetic biology and synthetic oil. It's considered a cool word. In the case of Algorithmic Intelligence, a nice touch is that the AI abbreviation would be a helpful carryover. The downside is that the word 'algorithmic' isn't any better than 'artificial' in the sense that, if anything, it seems less inspirational. There are a lot more of these kinds of candidates. We might consider Cognitive Computing (CC), which ditches the word 'intelligence' by using a reference to cognition instead. Does the proposed phrase seem rousing? Probably not. And so on. Shifting To Three Words Maybe the two-word sphere is limited and not conducive to our pursuit. Try these three-word-sized possibilities and see if any of them seem fitting: They each seem to be appealing and uplifting. That's good. A noteworthy concern is that they all include 'artificial intelligence' in the phrasing, which might create a bit of chaos. Will people be bewildered as to the difference between speaking of the two-word AI versus any of these three-word variations? Likely. You might know that there is already a movement afoot to switch over to three words. We seem to have gotten to describing AI as potentially becoming Artificial General Intelligence (AGI), which is AI on par with human intelligence, and there is Artificial Super Intelligence (ASI), denoting AI that is superhuman in intelligence. For my detailed explanation about AI, AGI, and ASI, see the link here. Will AGI and ASI ultimately replace the AI moniker? It's been slow going so far. Status Quo For Now The resistance to changing the AI moniker is bound to be extraordinarily sizable at this time. AI is on the upswing. Some would insist that if something isn't broken, there's no need to try and fix it. The same notion applies to the AI moniker. They would summarily reject the idea that the word 'artificial' needs to be replaced. Stick with what works. Whether this mood remains strident will soon be sorely tested. Here's what I mean. Suppose that all this talk and billions of dollars toward reaching AGI and ASI turn out to be a bust. Imagine that we continue to advance AI, but it doesn't attain AGI, nor does it achieve ASI. A lot of disappointment will ensue. There could be a 180-degree switcheroo that AI suddenly becomes a disappointment and no longer has the aura of greatness. In fact, the phrase might evoke downright fury and angst. Perhaps companies will rapidly drop the AI phrasing from their products and company websites. A bust in AI would be a downer, but you can bet that the pursuit of pinnacle AI would not stop. The dream of reaching pinnacle AI would remain. What might change is the naming of the field. No sense in letting the bitterness associated with a disappointing AI usurp the hope of attaining pinnacle AI. Slap on a new name and proceed forthwith. As Lord Byron so aptly stated: 'But words are things, and a small drop of ink, falling like dew, upon a thought, produces that which makes thousands, perhaps millions, think.' Depending on your belief in how contemporary AI is proceeding, it seems that the revered search for a suitable replacement phrase might indeed arise if the brass ring is not grasped. Time will tell.
Entrepreneur
an hour ago
- Entrepreneur
Defence Tech Startup Ammunic Systems Bags USD 1.1 Mn Funding
The funding will be deployed to boost research and development, scale manufacturing, expand the team, and enhance new and existing product lines focused on indigenous defence technology. You're reading Entrepreneur India, an international franchise of Entrepreneur Media. Ammunic Systems has raised USD 1.1 million in a seed funding round co-led by India Accelerator and Finvolve. The funding will be deployed to boost research and development, scale manufacturing, expand the team, and enhance new and existing product lines focused on indigenous defence technology. The capital infusion is set to support Ammunic Systems' ambition of accelerating its innovation roadmap and advancing India's self-reliance in strategic defence solutions. With a strong emphasis on cutting-edge technologies, the startup is positioning itself as a key player in modern warfare systems. "This investment from India Accelerator is not just capital; it is a vote of confidence in a mission we have poured our hearts into. At Ammunic Systems, we are not just building defence technologies; we are building belief — that India can lead from the front in securing its own future," said Priyanka Singhal, Founder of Ammunic Systems. Founded in 2024, Ammunic Systems is headquartered in Bengaluru. The company specialises in the research and development of advanced weapon systems. Its core expertise lies in electronic fuzes, warheads, and munition integration technologies. These products aim to offer precision, safety, and efficiency for current and future defence needs. Ammunic Systems provides end-to-end solutions across the defence product cycle — from concept design and testing to full-scale production. The startup aspires to become a leading innovator in the defence and aerospace sector by focusing on indigenous manufacturing and next-generation technologies. The investment community views such ventures as crucial to reducing India's dependency on defence imports. "India is at a critical point in its defence evolution, which makes indigenous technological capabilities national priorities. Ammunic Systems exemplifies the deep-tech innovation needed to strengthen our defence ecosystem," said Ashish Bhatia, Founder of India Accelerator and Co-founder of Finvolve.
