Microsoft recently announced that it would bring the beta of a gaming-specific version of Copilot to the Xbox platform in April. The new Copilot for Gaming is meant as a constant companion for anyone playing on Xbox, which can offer tips or hints on demand.

That leads me to ask: why would I want to use this at all? What’s the actual target audience here?

As I said recently on the GeekWire Podcast, machine learning as a whole is at that familiar tech stage where most of the new projects are solutions in search of a problem. If you look at recent product lineups from places like CES, advocates of the technology are still hung up on can instead of should.

Copilot for Gaming, as it was presented to me, falls squarely into that zone. It’s an attempt to improve the hobby by removing many of its human elements, so you never have to stop playing a game in order to look anything up. Your Copilot can ostensibly offer advice on how to start a new game, look up hints online, or provide personalized coaching.

The examples of use from the press briefing included asking Copilot for tips on what character to play in Blizzard’s hero shooter Overwatch 2 and how to get started as a brand new Minecraft player.

A Microsoft representative told me that Copilot for Gaming “accesses public sources of information using the Bing search index and results, and provides tailored responses for the individual player based on its understanding of the player’s activity and the games they’re playing on the Xbox platform.”

In addition, Microsoft currently plans that Copilot’s advice will source “the most accurate game knowledge,” including working directly with game studios.

Off the top, unless the plan is that every Xbox game will provide direct integration with material that was written in-house at Microsoft, Copilot for Gaming is still Copilot and will still have problems with its overall accuracy. As per a recent study by the Columbia Journalism Review, Copilot is wrong more often than it’s right.

The Microsoft representative was sure to note that “it’s important to check AI results against other trusted sources of information.” If that’s the case, then why am I using Copilot for Gaming at all, especially when it’s taking much-needed web traffic away from actual humans who want to provide me with the same guidance?

To be fair, the feature that Microsoft’s team initially led with struck me as being the most potentially useful. You can use Copilot for Gaming to give you a reminder of where you were and what you were doing the last time you played a particular game.

When I was a kid, one of my favorite games was 1993’s Phantasy Star IV, which has something like this as a menu feature, where you can trigger short conversations between your characters to remind you what you’re supposed to be doing. I’ve wanted to see more of that ever since. Many games are long and/or complicated enough that if you have to step away from them for a few days, you’ll have forgotten enough about the storyline or mechanics that you might as well start over. Having some kind of built-in refresher would be great.

The rest of Copilot for Gaming, however, strikes me as another example of an AI project with a reach that exceeds its grasp. Microsoft is asking players to trust an AI to provide them with accurate information at a point in time when AI-driven search is still distinctly flawed.

Full disclosure: I write strategy guides for sites like IGN, so I’m obviously biased. Even so, this whole project seems to be riding on the idea that players will happily sacrifice accuracy for convenience, and in my experience, such is not the case. The first time Copilot for Gaming gives someone incorrect gaming advice, it will take an unrecoverable PR hit.

Microsoft as a whole is smarter about the video game business than it often gets credit for being. Xbox may be stuck in a seemingly permanent third place position in the console market, but it’s got a few uncomplicated wins on its record, many of which revolve around its accessibility efforts.

The company recently announced the debut of its Xbox Adaptive Joystick, a new wired controller that’s intended for players with limited mobility, and which builds on its Adaptive Controller project.

You could also point to how the Xbox Game Pass is a great deal for budget-conscious players, particularly in conjunction with the Play Anywhere initiative, or Xbox’s history of outreach to the independent game developer community.

That’s part of what makes Copilot for Gaming frustrating. We’ve been on this train for a while, and we’re still regularly seeing AI researchers and developers trying to bring products to market that won’t work as advertised. The best case scenario for Copilot for Gaming is that it’s sort of helpful sometimes, at the price of removing yet more of the human element from the hobby. It’s a dumb move from a company that’s smarter than people think.

A.R. Moxon:

It’s best to understand that fascists see hypocrisy as a virtue. It’s how they signal that the things they are doing to people were never meant to be equally applied.

It’s not an inconsistency. It’s very consistent to the only true fascist value, which is domination.

It’s very important to understand, fascists don’t just see hypocrisy as a necessary evil or an unintended side-effect.

It’s the purpose. The ability to enjoy yourself the thing you’re able to deny others, because you dominate, is the whole point.

For fascists, hypocrisy is a great virtue — the greatest.

Yeah, this is basically why I don’t waste time anymore railing against the many hypocrisies of conservatives — they’re not gotchas that you’re catching them in, they’re part of the domination.

Tags: A.R. Moxon · politics

23andMe filed for Chapter 11 bankruptcy Sunday, leaving the fate of millions of people’s genetic information up in the air as the company deals with the legal and financial fallout of not properly protecting that genetic information in the first place. The filing shows how dangerous it is to provide your DNA directly to a large, for-profit commercial genetic database; 23andMe is now looking for a buyer to pull it out of bankruptcy.

23andMe said in court documents viewed by 404 Media that since hackers obtained personal data about seven million of its customers in October 2023, including, in some cases “health-related information based upon the user’s genetics,” it has faced “over 50 class action and state court lawsuits,” and that “approximately 35,000 claimants have initiated, filed, or threatened to commence arbitration claims against the company.” It is seeking bankruptcy protection in part to simplify the fallout of these legal cases, and because it believes it may not have money to pay for the potential damages associated with these cases. 

CEO and cofounder Anne Wojcicki announced she is leaving the company as part of this process. The company has the genetic data of more than 15 million customers.

According to its Chapter 11 filing, 23andMe owes money to a host of pharmaceutical companies, pharmacies, artificial intelligence companies (including a company called Aganitha AI and Coreweave), as well as health insurance companies and marketing companies. 

The filing is a devastating reminder that once you give your genetic information to a company like 23andMe, there is no way to have any clue what is going to happen to that data, how it is going to be analyzed, how it is going to be monetized, how it is going to be protected from hackers, and who it is going to be shared with for profit. Sharing your own DNA with 23andMe also necessarily implicates your close family members, who may or may not want their genetic information submitted to a company that is financially precarious and sitting on a trove of highly sensitive information.

On Friday, California Attorney General Rob Bonta issued an “urgent” alert to 23andMe customers telling them to ask the company to delete their data and destroy their genetic samples under a California privacy law: “Given 23andMe’s reported financial distress, I remind Californians to consider invoking their rights and directing 23andMe to delete their data and destroy any samples of genetic material held by the company.” 

Other genetic sequencing companies have shared customer information with police and governments, pharmaceutical companies, and health insurers. GED Match, a non-profit that once claimed it would protect customers’ genetic data, was sold to a for-profit company called Verogen, which works with the FBI and was later sold to a Dutch multinational conglomerate. Police now regularly attempt to identify suspects using information pulled from commercial genetic databases like the one that 23andMe has created.

23andMe’s bankruptcy means that the company will be put up for sale, and there’s no way of knowing who is going to buy it, why they will be interested, and what will become of its millions of customers’ DNA sequences. 23andMe has claimed over the years that it strongly resists law enforcement requests for information and that it takes customer security seriously. But the company has in recent years changed its terms of service, partnered with big pharmaceutical companies, and, of course, was hacked. 

In a letter to customers Sunday, 23andMe said “Your data remains protected. The Chapter 11 filing does not change how we store, manage, or protect customer data. Our users’ privacy and data are important considerations in any transaction, and we remain committed to our users’ privacy and to being transparent with our customers about how their data is managed.” It added that any buyer will have to “comply with applicable law with respect to the treatment of customer data,” which means essentially nothing because there are few laws that protect against the monetization of customer genetic data, as evidenced by the fact that other genetic databases proactively offer information to law enforcement and partner with big pharma. 

The company now could be sold to anyone, and there is no way to know what that buyer will want to do with the reams of genetic information it has collected. Customers, meanwhile, still have no way to change their underlying genetic data.

Tesla’s share price has been having a hard time of it lately. The stock has lost about half its value since its all-time high back in December, and, since Musk took office alongside Donald Trump in January, dropped for 7 consecutive weeks, rebounding only ever-so-slightly last week, after Musk got the president of the United States to turn the White House lawn into a cheesy Tesla (sorry, Tesler) dealership. Tesla stock dropped another 5 percent today, on a day when the overall market was slightly up.

I bookmarked this Bryce Elder column at the Financial Times back on January 31, and now seems like a good time to link to it:

The usual explanation for when Tesla trading resembles a Pump.fun shitcoin is: “because Elon talks a lot”. Here’s JPMorgan analyst Ryan Brinkman to expand on the theme:

It’s not clear to us why Tesla shares traded as much as +5% higher in the aftermarket Wednesday, although we have some leading theories. Perhaps it was management’s statement that it had identified an achievable path to becoming worth more than the world’s five most valuable companies taken together (i.e., more than the $14.8 trillion combined market capitalizations of Apple, Microsoft, NVIDIA, Amazon, & Alphabet). Or maybe it was management’s belief that just one of its products has by itself the potential to generate “north of $10 trillion in revenue”. It may have even related to management guidance for 2026 (no financial targets were provided, but it was said to be “epic”) and for 2027 and 2028 (“ridiculously good”).

Brinkman, who has a long-standing “underweight” rating on Tesla, is beginning to sound a bit exasperated:

[T]he company’s financial performance and Bloomberg consensus for revenue, margin, earnings, and cash flow all keep coming down, but analyst price targets and the company’s share price keep going up. For instance, Tesla has missed Bloomberg consensus EBIT in 9 of the past 10 quarters by an average of -16.3%.

Consistently missing estimates is one thing. What Tesla has been doing is consistently missing lowered estimates. [...]

Tesla’s biggest asset is hyperbole. The more extreme the hyperbole, the more valuable it gets. Maybe after-hours market participants understand the dynamics better than Tesla bears, so are primed to park fundamentals and trade on vibes. Or maybe something else entirely is going on.

Sounds a lot like the other guy at the White House Auto Mall.

Om Malik:

I have my own explanation, something my readers are familiar with, and it is the most obvious one. Just as Google is trapped in the 10-blue-link prison, which prevents it from doing something radical, Apple has its own golden handcuffs. It’s a company weighed down by its market capitalization and what stock market expects from it.

They lack the moral authority of Steve Jobs to defy the markets, streamline their product lineup, and focus the company. Instead, they do what a complex business often does: they do more. Could they have done a better job with iPadOS? Should Vision Pro receive more attention?

The answer to all those is yes. Apple has become a complex entity that can’t seem to ever have enough resources to provide the real Apple experience. What you get is “good enough.” And most of the time, I think it is enough — because what others have on the market is worse. They know how to build great hardware; it’s the software where they falter. In the case of Apple Intelligence, they have been caught short because others’ AI products, even when flawed, are significantly better than Apple’s own offerings.

Hardware inherently keeps a company honest in a way that software doesn’t. Hardware either works or it doesn’t. The only way to “upgrade” hardware is via installing newer software, or by taking the hardware apart and replacing physical components. It’s hard to think of a company, in any field, whose software is “better” than its hardware. Maybe Nintendo? But even with Nintendo, I’d say it’s more like their software is as good as their hardware. Also, an interesting thought that popped into my head reading Malik’s post just now: part of what makes Vision Pro so fascinating is that the software is better than the hardware. The hardware for immersive VR is so early-days that even the industry state-of-the-art — which is Vision Pro — stinks compared to where it’s going to be in even just five years. The 1984 Macintosh was a shitty computer with a 9-inch one-bit display, no hard drive, and an absurdly meager 128 kilobytes of RAM. But the software was amazing!

But the bigger, better point Malik makes is that “good enough” is enough to make Apple’s software seem ahead of its competition. I tried to make this point all the way back in 2007 with “Apple Needs a Nikon”, and I think the problem is worse now than it was then. No other company is even vaguely in Apple’s league. But Apple is sliding toward mediocrity on the software side. It very open for debate how far they’ve slipped. I, for one, would argue that they haven’t slipped far, and with an honest reckoning — especially with regard to everything related to Siri and AI — they can nip this in the bud. You might argue that they’ve slipped tremendously across the board. But what I don’t think is arguable is that their competition remains below Apple’s league. That’s what gives credence to the voices in Cupertino who are arguing that everything’s fine. Apple’s the only team in the top tier for UI design.

The best thing that could happen to Apple would be for Google to ship an Android Pixel experience that actually makes iPhone owners insanely jealous. Google is incapable of doing that through UI design. They’re incapable of catching up to Apple on hardware. But maybe on the AI front they can do it. Apple needs a rival.

I had the honor of another short segment on NPR’s Marketplace this morning. I spoke about the state of cyber crime, and the impact of US government changes on cyber defense.