Category: Hardware

I hope my headline is an overstatement, purely for job purposes, but in this week’s Vergecast artificial intelligence episode, we explore the world of large language models and how they might be used to produce AI-generated text in the future. Maybe it’ll give writers ideas for the next major franchise series, or write full blog posts, or, at the very least, fill up websites with copy that’s too arduous for humans to do.

Among the people we speak to is Nick Walton, the cofounder and CEO of Latitude, which makes the game AI Dungeon, which creates a plot in the game around what you put into it. (That’s how Walton ended up in a band of traveling goblins — you’ll just have to listen to understand how that makes sense!) We also chat with Samanyou Garg, founder of Writesonic, a company that offers various writing tools powered by AI. The company can even have AI write a blog post — I’m shaking! But really.

Anyway, toward the end of the episode, I chat with James Vincent, The Verge’s AI and machine learning senior reporter, who calms me down and helps me understand what the future of text-generation AI might be. He’s great. Check out the episode above, and make sure you subscribe to the Vergecast feed for one more episode of this AI miniseries, as well as the regular show. See you there!

Piaggio Fast Forward, a subsidiary of storied Italian automotive firm Piaggio, has launched its second robot, a compact version of its cargo-carrying bot Gita named Gitamini.

The form and function of Gitamini remain the same as that of full-sized Gita (the name is Italian for a small trip or outing). The robot consists of two large wheels, a central trunk, and a machine vision system that it uses to identify and follow its owner. Gitamini weighs 28 pounds and can carry up to 20 pounds in its interior for 21 miles. That makes an interesting comparison to Gita, which can carry more — 40 pounds but only for 12 miles.

Gitamini uses an array of cameras and sensors, including radar (not available for the original Gita), to navigate and follow its user. To activate this follow mode, you simply stand in front of the Gitamini and tap a pairing button. The robot will then lock on to you using vision only (no GPS or Bluetooth are utilized) and will follow you at speeds of up to 6mph.

Image: Piaggio Fast Forward

The robot’s trunk can be locked and its follow mode disabled, but there are no active theft mitigation features. When asked about this, Piaggio Fast Forward’s CEO Greg Lynn told The Verge that it was “unlikely someone could get away with walking away with it unnoticed” as it’s such a noticeable object. “A stolen Gita isn’t of much use to anyone as it uses a secure connection to a phone to be unlocked, updated, and used,” says Lynn. “We have yet to learn of a Gita being stolen or broken into while being used or when parked.”

The Gita has always been a bit of an odd product. It certainly looks fantastic, and videos suggest it works more or less as advertised (though it’s noisier than you might expect). But it’s not clear exactly who’s going to spend thousands of dollars on something that only carries a few bags and is stymied by steps and stairs. Gitamini doesn’t change any of these basic annoyances, though it is at least a little cheaper — it costs $1,850 (and will be available to buy from October 15th at mygita.com) while the launch sees the price of the original Gita drop to $2,950.

When we asked CEO Greg Lynn about the robot, he declined to share any sales figures with us but said there were Gita robots operating in “half the states in the US […] with a focus on the Southern belt where outdoor weather is more friendly year-round.”

“Most of the consumer Gitas are being used to replace car trips for neighborhood errands in a variety of communities, and they are used outdoors for round trips of a mile or more,” said Lynn. Though, he noted that the company had some business customers, too. There are currently Gitas in eight airports in the US (including JFK and LAX) and a number more in planned communities, like Water Street Tampa in Florida and Ontario Ranch in California.

AI research company OpenAI is releasing a new machine learning tool that translates the English language into code. The software is called Codex and is designed to speed up the work of professional programmers, as well as help amateurs get started coding.

In demos of Codex, OpenAI shows how the software can be used to build simple websites and rudimentary games using natural language, as well as translate between different programming languages and tackle data science queries. Users type English commands into the software, like “create a webpage with a menu on the side and title at the top,” and Codex translates this into code. The software is far from infallible and takes some patience to operate, but could prove invaluable in making coding faster and more accessible.

“We see this as a tool to multiply programmers,” OpenAI’s CTO and co-founder Greg Brockman told The Verge. “Programming has two parts to it: you have ‘think hard about a problem and try to understand it,’ and ‘map those small pieces to existing code, whether it’s a library, a function, or an API.’” The second part is tedious, he says, but it’s what Codex is best at. “It takes people who are already programmers and removes the drudge work.”

OpenAI used an earlier version of Codex to build a tool called Copilot for GitHub, a code repository owned by Microsoft, which is itself a close partner of OpenAI. Copilot is similar to the autocomplete tools found in Gmail, offering suggestions on how to finish lines of code as users type them out. OpenAI’s new version of Codex, though, is much more advanced and flexible, not just completing code, but creating it.

Codex is built on the top of GPT-3, OpenAI’s language generation model, which was trained on a sizable chunk of the internet, and as a result can generate and parse the written word in impressive ways. One application users found for GPT-3 was generating code, but Codex improves upon its predecessors’ abilities and is trained specifically on open-source code repositories scraped from the web.

This latter point has led many coders to complain that OpenAI is profiting unfairly from their work. OpenAI’s Copilot tool often suggests snippets of code written by others, for example, and the entire knowledge base of the program is ultimately derived from open-source work, shared to benefit individuals, not corporations. The same criticisms will likely be leveled against Codex, though OpenAI says its use of this data is legally protected under fair use.

When asked about these complaints, Brockman responds: “New technology is coming, we do need this debate, and there will be things we do that the community has great points on and we will take feedback and do things differently.” He argues, though, that the wider coding community will ultimately benefit from OpenAI’s work. “The real net effect is a lot of value for the ecosystem,” says Brockman. “At the end of the day, these types of technologies, I think, can reshape our economy and create a better world for all of us.”

Codex will also certainly create value for OpenAI and its investors. Although the company started life as a nonprofit lab in 2015, it switched to a “capped profit” model in 2019 to attract outside funding, and although Codex is initially being released as free API, OpenAI will start charging for access at some point in the future.

OpenAI says it doesn’t want to build its own tools using Codex, as it’s better placed to improve the core model. “We realized if we pursued any one of those, we would cut off any of our other routes,” says Brockman. “You can choose as a startup to be best at one thing. And for us, there’s no question that that’s making better versions of all these models.”

Of course, while Codex sounds extremely exciting, it’s difficult to judge the full scope of its capabilities before real programmers have got to grips with it. I’m no coder myself, but I did see Codex in action and have a few thoughts on the software.

OpenAI’s Brockman and Codex lead Wojciech Zaremba demonstrated the program to me online, using Codex to first create a simple website and then a rudimentary game. In the game demo, Brockman found a silhouette of a person on Google Images then told Codex to “add this image of a person from the page” before pasting in the URL. The silhouette appeared on-screen and Brockman then modified its size (“make the person a bit bigger”) before making it controllable (“now make it controllable with the left and right arrow keys”).

It all worked very smoothly. The figure started shuffling around the screen, but we soon ran into a problem: it kept disappearing off-screen. To stop this, Brockman gave the computer an additional instruction: “Constantly check if the person is off the page and put it back on the page if so.” This stopped it from moving out of sight, but I was curious how precise these instructions need to be. I suggested we try a different one: “Make sure the person can’t exit the page.” This worked, too, but for reasons neither Brockman nor Zaremba can explain, it also changed the width of the figure, squashing it flat on-screen.

“Sometimes it doesn’t quite know exactly what you’re asking,” laughs Brockman. He has a few more tries, then comes up with a command that works without this unwanted change. “So you had to think a little about what’s going on but not super deeply,” he says.

This is fine in our little demo, but it says a lot about the limitations of this sort of program. It’s not a magic genie that can read your brain, turning every command into flawless code — nor does OpenAI claim it is. Instead, it requires thought and a little trial and error to use. Codex won’t turn non-coders into expert programmers overnight, but it’s certainly much more accessible than any other programming language out there.

OpenAI is bullish about the potential of Codex to change programming and computing more generally. Brockman says it could help solve the programmer shortage in the US, while Zaremba sees it as the next step in the historical evolution of coding.

“What is happening with Codex has happened before a few times,” he says. In the early days of computing, programming was done by creating physical punch cards that had to be fed into machines, then people invented the first programming languages and began to refine these. “These programming languages, they started to resemble English, using vocabulary like ‘print’ or ‘exit’ and so more people became able to program.” The next part of this trajectory is doing away with specialized coding languages altogether and replacing it with English language commands.

“Each of these stages represents programming languages becoming more high level,” says Zaremba. “And we think Codex is bringing computers closer to humans, letting them speak English rather than machine code.” Codex itself can speak more than a dozen coding languages, including JavaScript, Go, Perl, PHP, Ruby, Swift, and TypeScript. It’s most proficient, though, in Python.

Codex also has the ability to control other programs. In one demo, Brockman shows how the software can be used to create a voice interface for Microsoft Word. Because Word has its own API, Codex can feed it instructions in code created from the user’s spoken commands. Brockman copies a poem into a Word document and then tells Word (via Codex) to first remove all the indentations, then number the lines, then count the frequency of certain words, and so on. It’s extremely fluid, though hard to tell how well it would work outside the confines of a pre-arranged demo.

If it succeeds, Codex might not only help programmers but become a new interface between users and computers. OpenAI says it’s tested Codex’s ability to control not only Word but other programs like Spotify and Google Calendar. And while the Word demo is just a proof of concept, says Brockman, Microsoft is apparently already interested in exploring the software’s possibility. “They’re very excited about the model in general and you should expect to see lots of Codex applications be created,” he says.

Amazon has updated the Alexa app on iOS so that you can access the voice assistant right from your home screen via a new widget. Everyone can use the assistant to remind specific members of your household to do tasks through a new “assign reminders” skill.

Due to the somewhat restrictive nature of the widgets on iOS, the new Ask Alexa widget isn’t so much Alexa itself as it is a link directly to the iOS app. But if you have the Alexa widget placed on any of your screens and you’ve already given the Alexa app permission to use your iPhone’s mic, you’ll be able to start making requests with a tap.

And now those requests can get a bit more granular. Amazon’s given Alexa the ability to assign reminders to specific members of your household if they have a Voice Profile set up on the same Amazon Alexa account. So if you say “Alexa, remind Jeff to take the lasagna out of the freezer at 10AM,” Alexa will be able to deliver the reminder to the right person, at the right time, through the Alexa app. You can add profiles to your Alexa account in Settings under Your Profile, and Amazon says you can assign relationship nicknames to each one, like mom, dad, daughter, etc.

Alexa picks up new features and skills on a monthly basis, but Amazon also announced plans in June to open up Alexa even further to third-party developers. Among many new APIs, developers will be able to create custom widgets for the Echo Show.

While there are big companies that do accept cryptocurrencies as payments, Amazon is not one of them, perhaps because of its unpredictable volatility. Yet the company is about to change its attitude towards cryptocurrencies and even plans to develop a special cryptocurrency and blockchain strategy. 

Business Insider has found an Amazon job listing that seeks a leader who will develop the retailer’s Digital Currency and Blockchain strategy as well as a product roadmap. The future employee of Amazon will be a part of The Amazon Payment Acceptance & Experience Team is responsible for ‘how Amazon’s customers pay on Amazon’s sites and through Amazon’s services around the globe,’ which pretty much implies that one of the world’s biggest retailers will start accepting cryptocurrency as payments sometimes in the future. 

In an interview with Gamingbolt – CEO and creative director at THE PARASIGHT, the developer behind Blacktail, gave some great insights about Microsoft’s next-gen budget offering, the Xbox Series S. Kapron believes that the Series S is an ambivalent piece of hardware.

The Xbox Series S is a drastically weaker console when compared to its bigger brothers such as the PS5 and Xbox Series X. In terms of raw TFLOPs, the Series S just has a third of the graphical grunt of the Series X. Thus, many seem to have doubts regarding its future, and whether it would be able to hold it’s own in next-gen titles. Kapron shares the sentiment to some extent, but also seems grateful that budget gamers can now get into next-gen gaming with the Series S.

“I think Series S is a very ambivalent piece of hardware. On the one hand, it makes the new generation much more affordable. On the other hand, everyone has doubts as to whether it won’t be a ball and chain, especially when the next gen will kick off for good. Personally, I think that despite the obvious difference in the target resolution in the future, we may also witness setting scaling between series X and S.”

While Microsoft’s claims of providing 1440p/60fps next-gen gaming looks to be an outstretched one, but the giant seems to be doubling down on making Series S as affordable as possible. In addition to providing gamers in select countries the option to purchase the console at a monthly payment with the Xbox All Access program, Xbox Game Pass ensures gamers get a healthy chunk of fresh offerings on a regular basis.

---

Restoration Hardware (RH) closed the most recent trading day at $685, moving +1.48% from the previous trading session. This change outpaced the S&P 500’s 1.02% gain on the day.

Heading into today, shares of the furniture and housewares company had lost 0.32% over the past month, lagging the Retail-Wholesale sector’s gain of 0.19% and the S&P 500’s gain of 3.01% in that time.

RH will be looking to display strength as it nears its next earnings release. On that day, RH is projected to report earnings of $6.45 per share, which would represent year-over-year growth of 31.63%. Meanwhile, our latest consensus estimate is calling for revenue of $972.26 million, up 37% from the prior-year quarter.

For the full year, our Zacks Consensus Estimates are projecting earnings of $22.62 per share and revenue of $3.68 billion, which would represent changes of +26.86% and +29.23%, respectively, from the prior year.

Any recent changes to analyst estimates for RH should also be noted by investors. These revisions typically reflect the latest short-term business trends, which can change frequently. With this in mind, we can consider positive estimate revisions a sign of optimism about the company’s business outlook.

Based on our research, we believe these estimate revisions are directly related to near-team stock moves. We developed the Zacks Rank to capitalize on this phenomenon. Our system takes these estimate changes into account and delivers a clear, actionable rating model.

The Zacks Rank system ranges from #1 (Strong Buy) to #5 (Strong Sell). It has a remarkable, outside-audited track record of success, with #1 stocks delivering an average annual return of +25% since 1988. Over the past month, the Zacks Consensus EPS estimate remained stagnant. RH currently has a Zacks Rank of #3 (Hold).

Looking at its valuation, RH is holding a Forward P/E ratio of 29.85. This represents a premium compared to its industry’s average Forward P/E of 15.56.

Story continues

Meanwhile, RH’s PEG ratio is currently 1.7. This metric is used similarly to the famous P/E ratio, but the PEG ratio also takes into account the stock’s expected earnings growth rate. The Retail – Home Furnishings industry currently had an average PEG ratio of 1.45 as of yesterday’s close.

The Retail – Home Furnishings industry is part of the Retail-Wholesale sector. This group has a Zacks Industry Rank of 18, putting it in the top 8% of all 250+ industries.

The Zacks Industry Rank gauges the strength of our individual industry groups by measuring the average Zacks Rank of the individual stocks within the groups. Our research shows that the top 50% rated industries outperform the bottom half by a factor of 2 to 1.

You can find more information on all of these metrics, and much more, on Zacks.com.

Want the latest recommendations from Zacks Investment Research? Today, you can download 7 Best Stocks for the Next 30 Days. Click to get this free report
 
RH (RH) : Free Stock Analysis Report
 
To read this article on Zacks.com click here.
 
Zacks Investment Research

If you follow technology news and policy at all, you may have heard about the right to repair — the idea that, by law or simply because it’s the right thing to do, companies that make products should provide the instructions so that people can repair and extend the life of their devices. What would the ideal device look like in that scenario?

It would probably be a lot like the Framework Laptop. The first device from Framework, the notebook (starting at $999 pre-configured or $749 for the barebones DIY Edition we tested). is  designed to be easily upgradeable, with the possibility of replacing the motherboard without tossing the whole laptop. It also allows for customizable side ports through a number of expansion cards that fit into the chassis. In theory, you’ll be able to consistently update this laptop rather than replace it entirely, reducing waste and getting precisely the laptop you want. It’s much easier to upgrade and fix than the best ultrabooks currently out there.

In my time with the DIY Edition (plus sampled components and expansion cards loaned by Framework), I was surprised at just how well this first-generation product seemed to come out. Yes, I have qualms with the reflective display and the plasticky trackpad. But I also got the motherboard out in less than 20 minutes.  While it’s promising that Framework is preparing to ship the first units (a hurdle that many companies haven’t passed), the company will really have to exist and thrive for a few years in order to see the Framework Laptop’s full potential. 

Design of the Framework Laptop 

On the outside, the Framework Laptop doesn’t look like anything special. Inside, it’s making a statement. In most of our reviews, we separate out the overall design of a notebook and how you upgrade it. But on the Framework Laptop, you can’t talk about one without the other. 

Atom Computing adds itself to a growing list of quantum systems makers with pedigreed founders, funding announcements, and a market that even the big players haven’t mastered. With no acquisition/cash-out goals apparent, no established market to chase, and competitive differentiation so nuanced, what’s the game?

If the last five years revolved around AI chip startups, expect the next five to shift to another upstart—the quantum device makers. If the neural network hardware startup crush taught us anything, it’s that it’s hard to challenge the largest chipmakers. We suspect a similar situation with the emerging quantum systems makers.

There are already existing large-scale players in the ecosystem (IBM, Google, and Microsoft) and more established startups, including D-Wave, are also worth mentioning. But even between all of these companies there is so little of practical, real-world value happening on these machines that they do not yet represent any upset to the traditional computing world.

Nonetheless, plenty will try. As with AI hardware, it’s easy to cloak real technical merit in magic-science speak and not have to explain technical differentiation. Without any benchmarks or even functional high-qubit devices that can operate at large enough scale to warrant a multi-vendor comparative effort beyond qubit count (itself not adequate in measuring performance) quantum startups can make almost any claim.

This is not to say the devices are invalid or worse, not even real/functional. It’s to say that it’s a tricky time to enter the quantum startup world in hardware. It’s not as simple as saying “it’s still too early” it’s that the market potential in the next five to ten years could still be minimal in reality. For those who do go the quantum route, how many companies are needed? And is it likely users will look to those with the most established, long-running quantum software stacks and hardware devices.

The quantum startups that cannot compete on time in the space, and who aren’t at liberty to/can’t explain what they do exactly, how it’s different from existing approaches, and how their software works in technical detail do have one last trick up their collective sleeves. Make waves by big, famous hires and raise a lot of capital on the power of those big, famous hires. That happened in the latter stages of the AI hardware game (and has in other areas in IT for years before that).

All of this was to introduce was a sideways way of introducing yet another quantum hardware maker into the space. There’s definitely some magic science speak here, and there’s definitely some funding and pedigree. But there are also a few things worth noting that take this company beyond a few others we’ve watched crop up with no real descriptions of what they do, how it’s different, how they make it, who will use it and how, etc.

This upstart is Atom Computing. Instead of calling a qubit a “qubit” they’re calling them “atoms”. They are one of several companies we’ll see in the coming year or two basing quantum systems on spin qubits. When we got the advance press release on Atom’s news that it’s raised $15 million, this sentence caught our eye: “Atom Computing is the first company to build nuclear-spin qubits out of an alkaline earth element.” We asked the startup’s CTO, Dr. Ben Bloom, what in the hell this means.

Our qubits are made of Strontium-87. There are more than 70 levels in our atom that have lifetimes of 10 seconds+. With our first-generation system, Phoenix, we’ve figured out how to control a subset of those levels that are intrinsically stable, made out of different configurations of the nucleus of an alkaline earth atom. This allows us to write quantum information into a scalable system that is shielded from the outside world, without having to resort to dilution refrigerators or other tricks.

So spin qubits. Got it.

From the press release (emphasis ours): “The company’s first-generation quantum computing system, Phoenix, is currently capable of trapping 100 atoms in a vacuum chamber with optical tweezers. Phoenix is able to rearrange and manipulate their quantum states with lasers. The system demonstrates exceptionally stable qubits at scale, with coherence times that are orders of magnitude greater than ever reported.”

When asked about whether there is room in the market for another quantum hardware maker, Atom Computing CEO, Rob Hays tells The Next Platform, “Even with incredible advances in computing performance in the exascale era, there are still mathematical problems, complex simulations, and AI models that still can’t be effectively solved with supercomputers alone. Quantum computers offer a new paradigm in computing that allow a massive continuum of solution space to be explored in parallel with a relatively small number of qubits and new quantum algorithms. We expect quantum computers and classical HPC clusters to be mated together to reach new heights in computing performance and solve these difficult problems together.”

Hays comes to the quantum startup world from the enterprise IT segment. Before Atom, Hays was Vice President and Chief Strategy Officer for Lenovo’s Infrastructure Solutions Group and spent 20 years at Intel, where he was Vice President and General Manager of Data Center Group Strategic Planning.CTO and co-founder Ben Bloom spent a year at another quantum hardware startup, Rigetti Computing and two years prior to that Intel as a module and yield integration engineer.

Atom Computing has been around for almost four years with Bloom as CEO until Hays stepped into the role this week. The company secured more than $15M in Series A funding which includes investment from Venrock, Innovation Endeavors, and Prelude Ventures. In addition, the National Science Foundation awarded the company three grants.

When asked how there is market room for another startup in the quantum systems space, Bloom tells us, “Atom Computing is dedicated to building useful, gate-based quantum computers, where every Atom equals 1 qubit. We believe the only way to build a scalable quantum computing system is to try new and exciting things. Our point-of-view is that long-lived, high-coherence, scalable systems are the only way to build a successful quantum computer. It’s about demonstrating performance at scale. We are committed to showcasing technical milestones and benchmarks that actually matter for creating a universal quantum computer.”

But here’s the question: what are the technical milestones and benchmarks that actually do matter in this nascent space?

In quantum at this moment, there is tremendous device diversity, but on the micro-level. There are differences in how qubits talk to one another, how tolerant to noise they are, how usable the software stack to interface with them has become and so on. Further, for people used to following systems, we have been trained to think in core counts and clock speeds. Qubit count doesn’t mean a thing if they can’t function together and a qubit, (or “atom”, or whatever you’d like to call it to make it sound different) so competing on that doesn’t work either.

Every quantum startup wants to come out of the gate looking different. It’s nearly impossible when even our smartest readers have a difficult time explaining in any level of technical detail what makes D-Wave’s approach different than IBM’s and so on. The opportunity for a marketing-driven startup to sweep in, blind VCs and the media with science-magic-talk and hyperbole is great.

Companies like Atom Computing and those who will surely follow with their own quantum hardware story are doing something difficult (stable spin qubits, for instance) have an equally tough challenge ahead: communicating past the first funding round about how, where, and why they’ll shave out any kind of market reach.

Unlike with the AI chip startups where it was clear in some cases companies were built for acquisition/cash-out, the big companies aren’t buying quantum startups. They have their own problems getting their own hardware/software stacks to work. So, again with the title: what are quantum hardware startups thinking?