Finding crypto’s floor

Crypto winter is in full swing. Layoffs at crypto firms, account freezes and plunging coin values show no signs of letting up.

Markets are being roiled this week. As the S&P 500 entered bear territory Monday, investors sold off their most speculative assets. That only increased crypto's freefall, and helped it lose its trillion-dollar market cap.

• After Celsius announced that it would stop withdrawals, swaps and transfers this weekend, Binance paused bitcoin withdrawals. The move was because of a “stuck transaction,” and withdrawals have since resumed, but it fueled investor uncertainty.
• The biggest crypto firm layoff so far happened Monday, when BlockFi announced that it’s cutting 20% of its staff.
• The crypto sell-off is intensifying, according to the MVIS CryptoCompare Digital Assets 100 Index, which measures the 100 top tokens.

What's different this time? Previous crypto crashes could be blamed on hackers, China’s ban on bitcoin and investor uncertainty. This year’s crash can be attributed to inflation, high interest rates and market uncertainty.

And experts expect the market to get worse before it gets better. Crypto exchange Luno's Vijay Ayyar thinks the price of bitcoin will continue to fall. But there may be hope on the horizon.

• Cubic Analytics’ Caleb Franzen said we can look to a Federal Reserve meeting tomorrow for a clue on what happens next. The Fed is expected to raise interest rates, which could ease inflation. But given crypto seems to behave a lot like stocks, that could be a further headwind for the digital currencies.
• “When we ask how long this crypto winter is going to last, it all comes down to the Fed,” Franzen said.

Crypto has crashed before, and it may crash again. But for now, maybe think twice before buying the dip.

The EU’s new tech crackdown

Social media companies aren’t moving fast enough to stamp out misinformation, according to the EU. Now those businesses have to detail their efforts or face massive fines.

The EU is taking big steps on misinformation.Facebook, Twitter, Google, Microsoft and TikTok have reportedly agreed to sign on to the bloc’s updated code on fighting misinfo, according to The Financial Times.

• The updated code, first introduced in 2018, forces platforms to reveal how exactly they are removing or preventing harmful content in the advertising and promoting of content.
• The code requires some serious concessions on tech giants’ part: It forces them to release country-by-country breakdowns of their efforts, not just global or Europe-focused data, a move that was previously resisted by big tech companies.
• The new rule will reportedly be enforced via Europe’s Digital Services Act. Companies that violate the rule will be fined in order to get them to comply.

The hope is for a more aggressive stance on misinformation from social media companies. Research shows that even as companies pledge to reduce misinfo, it still runs rampant.

• NYU researchers said Monday that YouTube has failed in its content moderation efforts, making it “a venue for political disinformation, public health myths, and incitement of violence.”
• Another recent report shows that content moderation efforts by social platforms haven’t stamped out the spread of climate change disinfo.

Experts say tech platforms need to be more accountable. The NYU researchers recommended radical transparency into how platforms are tackling disinformation. The EU is going to force the issue.

Why AI Twitter freaked out about 'sentient' AI

Twitter was ablaze this weekend, but it had nothing to do with Amber Heard. Instead, tech geeks were up in arms over a Googler’s contention that a natural-language-processing model developed by the company is “sentient,” and therefore may have achieved the AI holy grail: artificial general intelligence.

• Blake Lemoine, a Google engineer and philosophical rabble-rouser who was the subject of a Washington Post profile, thinks Google’s LaMDA model might have a mind of its own. Some people who read the article worried about the implications for humanity if AI becomes self-aware. Meanwhile, AI ethicists and pragmatists dismissed Lemoine’s contention.

“‘Sentient’ is being misapplied by many ML folks,” tweeted Oregon State University AI professor Thomas Dietterich. Author and NYU AI professor Gary Marcus agreed, noting, “[L]anguage these systems utter doesn’t actually mean anything at all. And it sure as hell doesn’t mean that these systems are sentient.”

• Meanwhile, some AI ethicists including former Googlers Timnit Gebru and Meredith Whittaker pointed to profit-driven AI business models as the key reason for the AGI hype.

While machine-learning models are probably nowhere near general intelligence, there is one thing Lemoine has pushed for that could matter sooner: the notion of legal personhood for LaMDA. If AI-based tech is deemed to be a person in the eyes of the law, it could have meaningful legal implications for liability, copyright and other issues.

This item first appeared in our Protocol Enterprise newsletter. Sign up here!

— Kate Kaye (email | twitter)

A MESSAGE FROM TRUSTED FUTURE

At the same time that the pandemic demonstrated all that is possible in an interconnected world, we saw in new and increasingly stark ways how certain communities continue to be marginalized and harmed by a persistent digital divide and how effectively that divide exacerbates our society’s other inequities.

Learn more

People are talking

John Oliver made a case for breaking up tech giants:

• “Ending a monopoly is almost always a good thing, whether it’s AT&T, or Standard Oil, or literally any game of Monopoly. When harmful monopolies end, innovation flourishes.”

Elon Musk told Tesla employees it’s been a rough quarter:

• “This has been a very tough quarter, primarily due to supply chain and production challenges in China. So we need to rally hard to recover!”

Making moves

Qualcomm bought Cellwize, a telecom software startup, for an undisclosed amount.

Boomi got a few new execs: PayPal’s Marcy Campbell joined as CRO; VMware’s Madeline Ling is the company’s new CFO; and Red Hat’s Fran Wilson joined as CMO.

Yoav Schwartzberg joined Melio as VP of Product. He previously spent a decade at Google, most recently as a director of Product Management.

Ira Wurcel is BitGo’s new general counsel. Wurcel’s been a lawyer for Deutsche Bank for over a decade.

Vincent Cirel is Gannett’s new CTO. Cirel was most recently the founder and managing director of Pivotal Technologies.

In other news

Horizon Worlds is expanding to the U.K. Meta also added a setting that turns a stranger's speech into gibberish. OK then!

Spotify formed a Safety Advisory Council to advise the company as it develops products and policies. What was that, did somebody mention Joe Rogan?

Elon Musk will talk to Twitter staff Thursday at a virtual all-hands meeting. It'll be the first time Musk has addressed staff since making his bid to buy the company.

SpaceX's Texas Starship base expansion plans got FAA approval. The company needs to address the environmental impact of its rocket launches in the area.

You know where there’s no lack of jobs? Among devs. This chart shows that the number of job openings for all kinds of devs is still on the up and up.

Microsoft's now legally required to remain neutral toward unions, according to a first-of-its-kind agreement between the company and the CWA.

BYOT means Bring Your Own Team. Hiring platform Manfred is testing out what it would be like for companies to hire entire teams.

Amazon’s actually going to do drone deliveries. Customers in Lockeford, California, will be able to get orders plopped in their backyard via Prime Air later this year.

How to create a masterpiece

OpenAI’s image generation tool DALL-E 2 is either the future of artificial intelligence or a great way to waste a few hours — or both. How it works: Enter keywords in a search bar as if you were Googling for something completely ridiculous (i.e. “Fred Durst eating Smartfood popcorn”). The tool spits out a series of images that are often delightful and at times unhinged using your keywords as a prompt. DALL-E (a mashup of Salvador Dalí and WALL-E) is a lot of fun and a useful tool for creative inspiration, and while the full experience is currently in private beta, you can check out a DALL-E mini version here to create your own masterpieces. (The mini version is not affiliated with OpenAI and is just an example of a similar AI-generated image creator — though it's equally fun!)

This story was updated June 14, 2022, to clarify information about DALL-E mini.

A MESSAGE FROM TRUSTED FUTURE

There is so much more we need to do to make sure our future is more equitable and inclusive and maximizes America’s potential. It is not enough just to ensure everyone is connected. We also need to extend the full scope of digital opportunity to the people, the communities, and the institutions.

Learn more