Tech Brew Ride Home - Wed. 06/12 – Elon Pulls His OpenAI Lawsuit
Episode Date: June 12, 2024Elon Musk has withdrawn his lawsuit against Sam Altman and OpenAI. I guess Wall Street likes Apple’s AI strategy. But why is Microsoft already putting the brakes on some of its AI features? BeReal g...ets acquired. What does the word “slop” mean when it comes to AI? And what happens when you add modern technology to the humble walkie talkie? Links: Elon Musk drops suit against OpenAI and Sam Altman (CNBC) X is officially making likes (mostly) private for everyone (Engadget) Apple quietly improves Mac virtualization in macOS 15 Sequoia (ArsTechnica) Microsoft is killing off GPT Builder in Copilot Pro for consumers, just three months after broad availability (XDA Developers) OpenAI ex-employees worry about company’s control over their millions of dollars in shares (CNBC) Photo-sharing app BeReal acquired by Voodoo for €500mn (FT) First Came ‘Spam.’ Now, With A.I., We’ve Got ‘Slop’ (NYTimes) Raleigh smart walkie-talkie startup Relay raises $35M from investors (Axios) Learn more about your ad choices. Visit megaphone.fm/adchoices
Transcript
Discussion (0)
On April 4th, 2023, around 2 in the morning, a man was found stabbed multiple times on a sidewalk in downtown San Francisco.
Hey, who did this to you?
What happened next turned the story into a political firestorm.
Reports have identified the victim as Bob Lee, the founder of Cash App.
From Bloomberg Podcasts, this is Foundering, the Killing of Bob Lee, beginning April 16.
Welcome to the Tech meme right home for Wednesday, June 12th, 2024. I'm Brian McCullough today.
Elon Musk has withdrawn his lawsuit against Sam Altman and Open AI. I guess Wall Street likes Apple's
AI strategy, but why is Microsoft already putting the brakes on some of its AI features?
Be Real gets acquired. What does the word slop mean when it comes to AI and what happens when you
add modern technology to the humble walkie-talkie? Here's what you miss today in the world of tech.
Elon Musk has withdrawn his lawsuit against OpenAI, Sam Altman, and Greg Brockman, which was filed in California back in February over breach of contract and fiduciary duty, quoting CNBC.
A hearing was scheduled for Wednesday in San Francisco, in which the judge was going to consider whether the case should be dismissed as requested by the defendants.
Experts told CNBC in March that the case was built on a questionable legal foundation because the contract at the heart of the suit was not a formal.
written agreement that was signed by all parties involved. Rather, Musk had alleged that the early
OpenAI team had set out to develop artificial general intelligence or AGI, quote, for the benefit of
humanity, but that the project has been transformed into a for-profit entity that's largely
controlled by principal shareholder Microsoft. Musk had used much of the 35-page complaint plus
attached exhibits. He filed in March to remind the world of his position in the creation of a company
that's since become one of the hottest startups on the planet. Musk's decision to file to dismiss
the suit came just one day after he publicly criticized OpenAI and its new partnership with Apple, end quote.
Meanwhile, X says the platform will make likes private for everyone this week, meaning users will no longer be able to see who liked someone else's post.
Quoting and Gadgett. Thanks to X showing what its users like on its platform, politicians and public personalities have been caught looking at salacious and unsavory tweets in the past.
Now the platform formerly known as Twitter is making likes mostly private and, according to,
to company chief Elon Musk. It's an important change so that people can quote,
like posts without getting attacked for doing so. The company originally launched the ability to
hide the likes tab as a perk for X premium subscribers last year. Keep spicy likes private, X said when it
announced the new feature. In a new tweet, X's engineering account has revealed that the social
network is making likes private for everyone this week. Users will no longer be able to see
who liked someone else's post, which means likes on the platform will no longer cause PR
crises for public figures who like sexual, hateful, and other unpalatable posts in general.
They can still see who liked their tweets, however, along with their like count and other metrics
for their own posts, end quote.
Apple's shares closed up 7.3% yesterday to $207.15 of share in their biggest one-day jump
since November 2022, giving the company a $3.18 trillion market cap just under.
Microsoft's $3.22 trillion market cap.
So investors liked what they saw at WWDC, I guess.
One other little nugget to come out from WWDC that I think developers would find pretty
important, quoting Ars Technica.
We've written before about Apple's handy virtualization framework in recent versions of Mac OS,
which allows users of Apple Silicon Macs with sufficient RAM, to easily set up MacOS and Linux
virtual machines using a number of lightweight third-party apps.
This is useful for anyone who needs to test software in multiple macOS versions but doesn't own a fleet of Mac hardware or multiple boot partitions.
Intel Macs support the virtualization framework too, but only for Linux VMs making it less useful.
But up until now, you haven't been able to sign into iCloud using macOS on a VM.
This made the feature less useful for developers or users hoping to test iCloud features in macOS
or whose apps rely on some kind of syncing with iCloud or people who just wanted easy access to their iCloud.
cloud data from within a VM. This limitation is going away in macOS 15 Sequoia, according to
developer documentation that Apple released yesterday, as long as your host operating system is
macOS 15 or newer, and your guest operating system is macOS 15 or newer. VMs will now be able
to sign into and use iCloud and other Apple ID-related services, just as they would when running
directly on the hardware. This is still limiting for developers who might want to run an older
version of macOS on their hardware while still testing macOS 15 in a VM or those who want to
do the reverse so that they can more easily support multiple versions of macOS with their apps.
It also doesn't apply to VMs that are upgraded from an older version of macOS to Sequoia.
It has to be a brand new VM created from a macOS 15 install image.
But it's a welcome change and it will steadily get more useful as Apple releases more macOS
versions in the future that can take advantage of it, end quote.
There is a developing narrative around AI, the idea that adding AI to every platform, do people actually want it?
Will people actually use it? Apple is about to find out, of course, but the developing narrative I'm talking about is that those who have gone first maybe aren't necessarily seeing untrammeled success.
Microsoft is removing the custom GPT builder feature from the consumer version of Copilot Pro on July 10th, just three months after.
after launching the feature. Quoting XDA developers. As discovered by Windows latest, Microsoft
began recently rolling out email notifications for GPT Builders' retirement. Although the email
doesn't go into a lot of details, it does offer ways to cancel the co-pilot subscription,
which may make sense for some, considering that the $20 per month subscription won't bundle
GPT Builder anymore. In a dedicated support document, Microsoft shed some light on the topic,
stating that it is evaluating its strategy regarding copilot extensibility in consumer use cases.
As such, it has now reached the conclusion that GPT builder for consumers is not a viable offering,
with the company still remaining committed to enterprise and commercial support for the technology.
Microsoft will remove GPT builder for consumers from July 10th, with associated GPTs and data being removed by July 14th.
Consumers cannot take any significant action in response to this retirement,
except save the instructions for their GPTs for reference or cancel their co-pilot pro subscriptions.
Microsoft will delete all co-pilot GPTs and related data from its platform,
and it will not be possible for customers to access the GPTs that they created previously.
The decision to kill off GPT Builders so soon after launch is rather surprising,
and it indicates that while the company is all in on its co-pilot and other AI initiatives,
it will only support them as long as it makes commercial sense to do so, end quote.
Add this to the developing narrative that there's a bit of turmoil among the rank and file over at OpenAI.
CNBC is reporting that current and former OpenAI staff are increasingly worried about its power over their equity,
with no IPO in sight, restrictive company policies, and more.
Quote, due to OpenAI skyrocketing valuation following the launch of Chad GPT in late 2022,
many early employees are sitting on millions of dollars worth of equity.
With no IPO on the horizon and a price tag that makes the company too expensive to be acquired,
The only way for shareholders to realize any value from their equity in the near term is through
secondary stock sales. However, current and former OpenAI employees have been increasingly
concerned about access to liquidity, according to interviews and documents shared internally.
Those fears have intensified in recent weeks. After reports that the company had the power
to clawback vested equity, said people familiar with the matter who asked not to be named
because the information they shared is confidential. In an attempt to assuage some of those concerns,
OpenAI recently circulated a document obtained by CNBC titled, Overview and Recap of OpenAI's
Tender Process, detailing how the company has conducted equity purchases in the past and how it plans
to handle them in the future. The issue has become a major topic of conversation at OpenAI
and among people who have recently left, according to internal documents, Slack messages,
and exit agreements viewed by CNBC, as well as conversations with multiple former OpenAI employees.
Open AI has told employees that it will try to hold one tender offer roughly every year,
But that depends on how both the company and the market are faring at the time, a person with
knowledge of the matter said. As OpenAI has grown, the company has used aggressive tactics
to get employees to sign exit agreements that affect the future of their stock holdings.
If you have any vested units and you do not sign the exit documents, including the general
release as required by company policy, it is important to understand that, among other things,
you will not be eligible to participate in future tender events or other liquidity opportunities
that we may sponsor or facilitate as a private company, OpenAI wrote in the agreement which was
viewed by CNBC. The paperwork for departing employees says that in order to participate in tender
events and liquidity opportunities, the person must be in compliance with, quote, all applicable
company policies as determined by OpenAI. Last month, OpenAI announced it would
backtrack on a controversial decision to make former employees choose between signing a non-disparagement
agreement that would never expire and keeping their vested equity in the company. An internal memo
viewed by CNBC was sent to former employees and shared with current staffers, end quote.
Video game and app developer Voodoo has acquired a photo sharing app B-Real and a deal valuing B-Reel at
500 million euros. Voodoo's Ameri Rofi will become B-Reel's CEO, quoting the Financial Times.
Voodoo plans to launch paid advertising on B-Reel in an effort to move the company towards profitability.
Its chief executive, Alexandre Yadze, told the Financial Times,
adding that it wanted to expand the company's active user base. The takeover was advised by J.P.
Morgan and is the biggest in Voodoo's history. Ameri Raffee, head of Voodoo-owned social media app
Wiz, will become chief executive of Be Real, while Alexis Berriott, who founded the platform in
2019, age 26, will remain, quote, actively involved. Jean de la Rochevard, a Be Real board member
and partner at Kima Ventures, an investment company co-founded by French billionaire Zavier Neal,
called the sale a, quote, wise decision, especially as big financing rounds had become rarer,
while investors focused on the much hyped artificial intelligence sector. Be Real promotes authenticity
on its platform, in contrast to the filters and editing commonplace on rivals such as Snapchat
and Instagram. The Paris-based tech company, however, has largely stagnated in terms of user growth
since 2022 when its popularity surged among teenagers and young adults. Be Real has in the past been
financed through venture capital, raising a total of $90 million, according to data.
provider crunch base. B-REL's monthly active users have held steady at about 50 million since the end of
2022, according to data from Market Intelligence Company Sensor Tower. Downloads, however, tumbled from a
peak of about $35 million in the third quarter of 2022 to $5 million in the three months to March this
year. Although the growth has slowed significantly, Yadzi said the majority of B-reels users
still logged in at least six days a week, making it a, quote, impressive prospect for expansion and
monetization, end quote.
This is news you can use, I guess, knowledge you can use to keep cutting edge, keep hip,
gluing you into the slang the kids are starting to use so you can be in the know.
The New York Times took a look at how some people are using the term slop as a descriptor for low-grade AI material.
Quote, the use of slop as a descriptor for low-grade AI material seemingly came about in reaction to the release of AI art generators in 2022.
Some have identified Simon Willison, a developer as an early adopter of the term, but
Mr. Wilson, who has pushed for the phrase's adoption, said it was in use long before he found it.
I think I might actually have been quite late to the party, he said in an email.
The term has sprung up in 4chan, Hacker News, and YouTube comments, where anonymous
posters sometimes project their proficiency in complex subject matter by using in-group language.
What we always see with any slang is that it starts in a niche community and then spreads from there.
Mr. Aleski said, usually coolness is a factor that helps it spread.
but not necessarily. Like, we've had a lot of words spread from a bunch of coding nerds, right?
Look at the word spam. Usually the word is created because there is a particular group with shared
interests with a shared need to invent words, end quote. With the dominant search engines,
having made AI a priority, it appears that vast quantities of information generated by machines,
rather than largely curated by humans, will be served up as a daily part of life on the
internet for the foreseeable future. Hence the term slop, which conjures images of heaps of unappetizing
food being shoveled into troughs for livestock. Like that type of slop, AI-assisted search comes together
quickly, but not necessarily in a way that critical thinkers can stomach. Kristen Hammond,
the director of Northwestern University Center for Advanced Safety of Machine Intelligence,
noted a problem in the current model. The information from AI overview is being presented as a
definitive answer rather than as a place to start an internet user's research into a given subject.
You search for something and you get back what you need in order to think, and it actually
encourages you to think, Mr. Hammond said. What it's becoming in this integration with large language
models is something that does not encourage you to think. It encourages you to accept, and that I think is
dangerous. As people become accustomed to AI's increasing role in how the internet works, Mr. Willison,
who identified himself as an optimist for AI when it is used correctly, thought that slop could
become the go-to term for the lesser forms of machine-generated content. Society needs concise ways to
talk about modern AI, both the positives and the negatives, he said.
said, ignore that email, it's spam, and ignore that article, its slop are both useful lessons, end quote.
Finally today, interesting gadget, interesting raise, Raleigh, North Carolina-based relay,
which makes walkie-talkies that can provide real-time location data, do live translation, and more,
raised a $35 million series B led by G2, quoting Axios.
Relay makes what is essentially a smart walkie-talkie that enables managers to communicate to workers remotely,
Because it is connected to the internet, it can also provide real-time location data and insights.
The tool has proven useful in fields like hospitality with hotel workers using them between floors,
as well as in healthcare, education, and manufacturing, Chung said.
Relay's device, which include smartphone app integrations, can also do real-time language
translation between workers and serve as a safety alert if a team member needs immediate assistance.
David Mount, a partner at G2, said that his firm has spent the past two years researching relay
and came away impressed by how efficiently the company had grown its customer base.
Venture capitalists can sometimes be super focused on what's possible in 10 years, he told Axios,
but as we started talking to customers, we learned in a lot of places Relay is replacing
walkie-talkies, and walkie-talkies have not been innovated around in a long time.
So just being a better, less expensive and way more feature-rich walkie-tucky is a massive market,
he added.
And then it also provides a tool that over time we'll be able to do all this other stuff around
data collection. The company founded in 2015 and headquartered in North Hills, declined to share
valuation or revenue numbers with the raise, but Chung said that its revenues grew by
120% in 2023, end quote. Nothing more for you today. Talk to you tomorrow.
