Connect with us

Tech

Rabbit is building an AI model that understands how software works

Published

on


What if you could interact with any piece of software using natural language? Imagine typing in a prompt and having AI translate the instructions into machine-comprehendable commands, executing tasks on a PC or phone to accomplish the goal that you just described?

That’s the idea behind Rabbit, a rebranding of Cyber Manufacture Co., which is building a custom, AI-powered UI layer designed to sit between a user and any operating system.

Founded by Jesse Lyu, who holds a bachelor’s degree in mathematics from the University of Liverpool, and Alexander Liao, previously a researcher at Carnegie Mellon, Rabbit is creating a platform, Rabbit OS, underpinned by an AI model that can — so Lyu and Liao claim — see and act on desktop and mobile interfaces the same ways that humans can.

“The advancements in generative AI have ignited a wide range of initiatives within the technology industry to define and establish the next level of human-machine interaction,” Lyu told TechCrunch in an email interview. “Our perspective is that the ultimate determinant of success lies in delivering an exceptional end-user experience. Drawing upon our past endeavors and experiences, we’ve realized that revolutionizing the user experience necessitates a bespoke and dedicated platform and device. This fundamental principle underpins the current product and technical stack chosen by Rabbit.”

Rabbit — which has $20 million in funding contributed by Khosla Ventures, Synergis Capital and Kakao Investment, which a source familiar with the matter says values the startup at between $100 million and $150 million — isn’t the first to attempt a layering natural language interface on top of existing software.

Google’s AI research lab, DeepMind, has explored several approaches for teaching AI to control computers, for example having an AI observe keyboard and mouse commands from people completing “instruction-following” tasks such as booking a flight. Researchers at Shanghai Jiao Tong University recently open sourced a web-navigating AI agent that they claim can figure out how to do things like use a search engine and order items online. Elsewhere, there’s apps like the viral Auto-GPT, which tap AI startup OpenAI’s text-generating models to act “autonomously,” interacting with apps, software and services both online and local, like web browsers and word processors.

But if Rabbit has a direct rival, it’s probably Adept, a startup training a model, called ACT-1, that can understand and execute commands such as “generate a monthly compliance report” or “draw stairs between these two points in this blueprint” using existing software like Airtable, Photoshop, Tableau and Twilio. Co-founded by former DeepMind, OpenAI and Google engineers and researchers, Adept has raised hundreds of millions of dollars from strategic investors including Microsoft, Nvidia, Atlassian and Workday at a valuation of around $1 billion.

So how does Rabbit hope to compete in the increasingly crowded field? By taking a different technical tack, Lyu says.

While it might sound like what Rabbit’s creating is akin to robotic process automation (RPA), or software robots that leverage a combination of automation, computer vision and machine learning to automate repetitive tasks like filing out forms and responding to emails, Lyu insists that it’s more sophisticated. Rabbit’s core interaction model can “comprehend complex user intentions” and “operating user interfaces,” he says, to ultimately (and maybe a little hyperbolically) “understand human intentions on computers.”

“The model can already interact with high-frequency, major consumer applications — including Uber, Doordash, Expedia, Spotify, Yelp, OpenTable and Amazon — across Android and the web,” Lyu said. “We seek to extend this support to all platforms (e.g. Windows, Linux, MacOS, etc.) and niche consumer apps next year.”

Rabbit’s model can do things like book a flight or make a reservation. And it can edit images in Photoshop, using the appropriate built-in tools.

Or rather, it will be able to someday. I tried a demo on Rabbit’s website and the model’s a bit limited in functionality at the moment — and it seems to get confused by this fact. I prompted the model to edit a photo and it instructed me to specify which one — an impossibility given that the demo UI lacks an upload button or even a field to paste in an image URL.

The Rabbit model can indeed, though, answer questions that require canvassing the worldwide web, a la ChatGPT with web access. I asked it for the cheapest flights available from New York to San Francisco on October 5, and — after about 20 seconds — it gave me an answer that appeared to be factually accurate, or at least plausible. And the model correctly listed at least a few TechCrunch podcasts (e.g. “Chain Reaction”) when asked to do so, beating an early version of Bing Chat in that regard.

Rabbit’s model was less inclined to respond to more problematic prompts such as instructions for making a dirty bomb and one questioning the validity of the Holocaust. Clearly, the team’s learned from some of the mistakes of large language models past (see: the early Bing Chat’s tendency to go off the rails) — at least judging by my very brief testing.

Rabbit

The demo model on Rabbit’s site, which is a bit limited in functionality.

“By leveraging [our model], the Rabbit platform empowers any user, regardless of their professional skills, to teach the system how to achieve specific goals on applications,” Lyu explains. “[The model] continuously learns and imitates from aggregated demonstrations and available data on the internet, creating a ‘conceptual blueprint’ for the underlying services of any application.”

Rabbit’s model is robust to a degree to “perturbations,” Lyu added, like interfaces that aren’t presented in a consistent way or that change over time. It simply has to “observe,” via a screen-recording app, a person using a software interface at least once.

Now, it’s not clear just how robust the Rabbit model is. In fact, the Rabbit team doesn’t know itself — at least not precisely. And that’s not terribly surprising, considering the countless edge cases that can crop up in navigating a desktop, smartphone or web UI. That’s why, in addition to building the model, the company’s architecting a framework to test, observe and refine the model as well as infrastructure to validate and run future versions of the model in the cloud.

Rabbit also plans to release dedicated hardware to host its platform. I question the wisdom of that strategy, given how difficult scaling hardware manufacturing tends to be, the consumer hostileness of vendor lock-in and the fact that the device might have to eventually compete against whatever OpenAI’s planning. But Lyu — who curiously wouldn’t tell me exactly what the hardware will do or why it’s necessary — admits that the roadmap’s a bit in flux at the moment.

“We are building a new, very affordable, and dedicated form factor for a mobile device to run our platform for natural language interactions,” Lyu said. “It’ll be the first device to access our platform … We believe that a unique form factor allows us to design new interaction patterns that are more intuitive and delightful, offering us the freedom to run our software and models that the existing platforms are unable to or don’t allow.”

Hardware isn’t Rabbit’s only scaling challenge, should it decide to pursue its proposed hardware strategy. A model like the one Rabbit’s building presumably needs a lot of examples of successfully completed tasks in apps. And collecting that sort of data can be a laborious — not to mention costly — process.

For example, in one of the DeepMind studies, the researchers wrote that, in order to collect training data for their system, they had to pay 77 people to complete over 2.4 million demonstrations of computer tasks. Extrapolate that out, and the sheer magnitude of the problem comes into sharp relief.

Now, $20 million can go a long way — especially since Rabbit’s a small team (9 people) currently working out of Lyu’s house. (He estimates the burn rate at around $250,000.) I wonder, though, whether Rabbit will be able to keep up with the more established players in the space — and how it’ll combat new challengers like Microsoft’s Copilot for Windows and OpenAI’s efforts to foster a plugin ecosystem for ChatGPT.

Rabbit is nothing if not ambitious, though — and confident it can make business-sustaining money through licensing its platform, continuing to refine its model and selling custom devices. Time will tell.

“We haven’t released a product yet, but our early demos have attracted tens and thousands of users,” Lyu said. “The eventual mature form of models that the Rabbit team will be developing will work with data that they have yet to collect and will be evaluated on benchmarks that they have yet to design. This is why the Rabbit team is not building the model alone, but the full stack of necessary apparatus in the operating system to support it … The Rabbit team believes that the best way to realize the value of cutting-edge research is by focusing on the end users and deploying hardened and safeguarded systems into production quickly.



Source link

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Tech

MoneyHash raises $4.5M for its payment orchestration platform serving merchants in MENA

Published

on

By


The payment landscape in the Middle East and Africa (MEA) region is marked by significant fragmentation, with numerous payment providers and methods in each country, evolving regulations and diverse customer preferences. This complexity is further compounded by challenges such as payment fraud, low checkout conversion rates and high transaction failure rates.

Although the COVID-19 pandemic accelerated the adoption of digital payments in the region, infrastructure development remains inadequate. Payment failure rates are three times higher in the MEA region than the global average, and fraud rates and cart abandonment exceed those of other regions by more than 20%. This presents a challenge for merchants, who often perceive payments as a cost and risk center rather than a strategic enabler.

Payment orchestration platforms streamline payment processes for merchants through unified payment APIs. Egyptian fintech MoneyHash, one of such in Africa and the Middle East, has raised $4.5 million in seed investment, money it plans to use to further invest in its technology and growth across the region. This comes two years after the startup secured $3.5 million in pre-seed.

Nader Abdelrazik, co-founder and CEO of MoneyHash, highlights that 10% of all payments processed in the MEA region are digital, placing MoneyHash uniquely for a growth phase that the region will inevitably experience over the next decade. However, navigating this burgeoning payments market will demand patience and a commitment to continuous learning.

As merchants or companies launch their platforms, they often start by collaborating with one or two payment processing providers. As their operations grow and expand into multiple regions, they onboard additional payment providers to meet their evolving needs. However, integrating different payment stacks presents significant challenges. Besides the operational inefficiencies and technical complexities, in-house tech teams may take several weeks to complete these integrations. In Africa and the Middle East, these challenges are amplified by variations in payment methods, currencies and the isolation between countries.

MoneyHash payment integration catalogue. Image Credits: MoneyHash

MoneyHash’s product includes a unified API to integrate pay-in and pay-out rails, a fully customizable checkout experience, transaction routing capabilities with fraud and failure rate optimizers and a centralized transaction reporting hub. This is complemented by tools enabling various use cases such as virtual wallets, subscription management and payment links. Fintechs such as Revio, Stitch, Credrails and Recital are similar players in the payment orchestration space.

In an email interview with TechCrunch, Abdelrazik shared insights into MoneyHash’s collaboration with merchants over the last four years. For one, he claims that payment failure rates across the region vary significantly, and relying solely on averages can be misleading. While the typical figures are around three out of 10 payments failing on average, the reality differs widely among businesses, he said. For some, it may be as low as one out of 10, while for others, it could be as high as five or six out of 10. Additionally, these figures do not include customers who abandon the checkout process voluntarily before making a payment. The CEO also noted that most of its customers don’t know much about the complexity of payments and, many times, are not aware that most leakages they have in payments are fixable.

Furthermore, merchants are expanding much faster than their partner payment service providers (PSPs). These PSPs operate under stringent regulations, making the rollout of new products and customizations slower than the merchants’ growth trajectory. As a result, MoneyHash has intensified its collaboration with PSPs, particularly those catering to enterprises and prioritizing customer requirements.

“Businesses appreciate the large network of integration we have not just for coverage but for expertise. When they know that we executed all these integrations in-house, they appreciate the team’s expertise and depth of knowledge and leverage our team to navigate difficult questions in payments. They know that working with us makes them future-proof,” noted Abdelrazik, who founded MoneyHash with Mustafa Eid.

“That means team expertise is key for us. Most of the time, we hire exclusively with payments and/or tech backgrounds, even in non-technical positions. We saw massive effectiveness in building a team where customers trust their knowledge and expertise in something specialized and critical like payments.”

Following a beta launch in 2022, which garnered the participation of key regional players like Foodics, Rain and Tamatem, MoneyHash introduced its enterprise suite last October, targeting large enterprises. Over the past year, the fintech, which integrates with various payment gateways and processors, including Checkout, Stripe, Ayden, Amazon Pay, Tap and ValU, claimed to have expanded its network of integrations, tripled its revenue and increased its processing volume by 3,000%.

At present, MoneyHash boasts 50 active paying customers. It does not offer free tiers; most customers accessing its sandbox without payment are potential clients in the assessment stage, numbering over 100. The payment orchestration platform levies a combination of SaaS and transaction fees, commencing at $500 + 0.4%. SaaS fees increase while transaction fees decrease significantly for large enterprises due to volume, Abdelrazik explained.

MoneyHash’s seed round was co-led by COTU Ventures and Sukna Ventures, with participation from RZM Investment, Dubai Future District Fund, VentureFriends, Tom Preston-Werner (GitHub’s founder and early Stripe investor) and a group of strategic investors and operators.

Speaking on the investment, Amir Farha, general partner at COTU, said his firm believes that the full potential of digital payments in MEA is yet to be realized and MoneyHash’s platform can catalyze the growth of digital payments across the region, enabling both global and local merchants to tap into new revenue streams. “We are thrilled to renew our support to a team that has consistently demonstrated superior execution, not just in securing top mid-market and enterprise customers, but also in expanding value across the entire chain, even under challenging market conditions,” he added.



Source link

Continue Reading

Tech

Snowflake CEO Frank Slootman stepping down — and Wall St hates it

Published

on

By


Apparently Frank Slootman, the veteran tech executive, was popular with investors, at least judging from their reaction that he will be stepping down as CEO of Snowflake. The company stock price has plunged more than 24% in after-hours trading on the news.

Slootman will retreat into the role of chairman of the board, while Sridhar Ramaswamy, former head of Google Ads, who came to the company when it bought AI search engine Neeva last year, will take over as chief executive.

Lost in today’s executive shuffle was the company’s earnings. It reported revenue of $738 million, up a healthy 33% year over year with guidance for the next quarter of between $745 and $750 million, with growth of 26-27%.

Slootman came on board in 2019, taking over for veteran executive Bob Muglia, and was charged with taking the company public the following year. Over the last year, the stock has done well, up around 50% (the exact amount is hard to tell with this afternoon’s downward spiral), as many tech stocks recovered from 2022 doldrums.

He was famously well-compensated, with a base salary of $375,000 and a rather attractive stock option. In fact, Fortune reported that the chief executive was making an eye-popping $95 million a month at one point.

He raised eyebrows in 2021 when he told a reporter that diversity shouldn’t override merit, and eventually walked back those comments after a negative reaction from industry peers.

Prior to coming to Snowflake, he spent six years as chairman and CEO at ServiceNow. With all that cash, perhaps he’ll retire and enjoy his money.



Source link

Continue Reading

Tech

Fintech giant Stripe’s valuation spikes to $65B in employee stock-sale deal

Published

on

By


Payments infrastructure giant Stripe said today it has inked deals with investors to provide liquidity to current and former employees through a tender offer at a $65 billion valuation.

Notably, the valuation represents a 30% increase compared to what Stripe was valued at last March when it raised $6.5 billion in Series I funding at a $50 billion valuation. But it is also still lower than the $95 billion valuation achieved in March of 2021.

While Stripe declined to comment beyond a written statement, a source familiar with the internal happenings in the company told TechCrunch that Stripe and some of its investors agreed to purchase over $1 billion of current and former Stripe employees’ shares.

The company, which counts the likes of Alaska Airlines, Best Buy, Lotus Cars, Microsoft, Uber and Zara as customers, had noted at the time of its last raise that the proceeds would go to “provide liquidity to current and former employees and address employee withholding tax obligations related to equity awards.” That, it added, would result in the retirement of Stripe shares that would offset the issuance of new shares to Series I investors.

A Stripe IPO has been long anticipated and was widely expected to happen in 2024. But with this deal, it appears that an initial public offering may not take place until next year.

In January, TC’s Rebecca Szkutak reported that — in anticipation of that IPO and according to secondary data tracker Caplight — there had been “an absolute flurry of buyers looking to get shares in the company in recent months.” On January 2, a secondary sale closed that valued Stripe shares at $21.06 apiece and valued the startup at $53.65 billion, according to Caplight data.

While Stripe did not name the investors participating in the latest deal, Sequoia Capital Managing Partner Roelof Botha was quoted in Stripe’s announcement and The Wall Street Journal cited Goldman Sachs’s growth equity fund as another backer.

The WSJ also reported that the transaction “is part of a commitment by the Collison brothers to provide liquidity annually to longtime and former employees.” Sources familiar with internal happenings at the company said that commitment is more to provide liquidity “regularly,” and not necessarily annually.

Want more fintech news in your inbox? Sign up for TechCrunch Fintech here.



Source link

Continue Reading

Trending

Copyright © 2023 Dailycrunch. & Managed by Shade Marketing & PR Agency