Start Up No.1935: Google calls in Page and Brin over ChatGPT, CNet pauses AI writers, Amazon cuts drone team, teeny Twitter, and more

The latest job cuts at Microsoft are hitting the HoloLens team. Strange how the mixed-reality headset hasn’t caught on. CC-licensed photo by sndrv on Flickr.

You can sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.

A selection of 11 links for you. Use them wisely. I’m @charlesarthur on Twitter. Observations and links welcome.

Google calls in Larry Page and Sergey Brin to tackle ChatGPT and AI chatbots • The New York Times

Nico Grant:


Last month, Larry Page and Sergey Brin, Google’s founders, held several meetings with company executives. The topic: a rival’s new chatbot, a clever A.I. product that looked as if it could be the first notable threat in decades to Google’s $149bn search business.

Mr. Page and Mr. Brin, who had not spent much time at Google since they left their daily roles with the company in 2019, reviewed Google’s artificial intelligence product strategy, according to two people with knowledge of the meetings who were not allowed to discuss them. They approved plans and pitched ideas to put more chatbot features into Google’s search engine. And they offered advice to company leaders, who have put A.I. front and center in their plans.

The re-engagement of Google’s founders, at the invitation of the company’s current chief executive, Sundar Pichai, emphasized the urgency felt among many Google executives about artificial intelligence and that chatbot, ChatGPT.

…The new A.I. technology has shaken Google out of its routine. Mr. Pichai declared a “code red,” upending existing plans and jump-starting A.I. development. Google now intends to unveil more than 20 new products and demonstrate a version of its search engine with chatbot features this year, according to a slide presentation reviewed by The New York Times and two people with knowledge of the plans who were not authorized to discuss them.

…“This is a moment of significant vulnerability for Google,” said D. Sivakumar, a former Google research director who helped found a start-up called Tonita, which makes search technology for e-commerce companies. “ChatGPT has put a stake in the ground, saying, ‘Here’s what a compelling new search experience could look like.’”


If Chat GPT were actually accurate, sure, it could be a search experience. But it’s not. However the fact that Page and Brin have dragged themselves back into the office is very notable indeed. Google is worried.
unique link to this extract

CNET pauses publishing AI-written stories after disclosure controversy • The Verge

Mia Sato:


CNET will pause publication of stories generated using artificial intelligence “for now,” the site’s leadership told employees on a staff call Friday.

The call, which lasted under an hour, was held a week after CNET came under fire for its use of AI tools on stories and one day after The Verge reported that AI tools had been in use for months, with little transparency to readers or staff. CNET hadn’t formally announced the use of AI until readers noticed a small disclosure.

“We didn’t do it in secret,” CNET editor-in-chief Connie Guglielmo told the group. “We did it quietly.”

CNET, owned by private equity firm Red Ventures, is among several websites that have been publishing articles written using AI. Other sites like Bankrate and would also pause AI stories, executives on the call said.

Futurism noted that CNET and Bankrate appeared to have stopped running AI stories as early as Wednesday.

…The AI, which is as of yet unnamed, is a proprietary tool built by Red Ventures, according to Davis. AI editors are able to choose domains and domain-level sections from which to pull data from and generate stories; editors can also use a combination of AI-generated text and their own writing or reporting.

Turrentine declined to answer staff questions about the dataset used to train AI in today’s meeting as well as around plagiarism concerns but said more information would be available next week and that some staff would get a preview of the tool.

Leadership also differentiated between the unnamed internal tool and other automated technology Red Ventures uses on its sites to auto-insert numbers into mortgage rate and refinance rate stories, which The Verge reported had been in use for far longer but that the company didn’t disclose.


I love “we didn’t do it in secret, we did it quietly.” One for the ages. I suspect that the private equity company will be restarting the AI writing in se.. quietly, and fairly soon. When did a private equity company ever turn down a chance to cut costs?
unique link to this extract

Amazon drone unit hit with layoffs as long-awaited program launches • CNBC

Annie Palmer:


In 2013, Amazon founder Jeff Bezos appeared on CBS’s “60 Minutes” to reveal a futuristic plan his company had been secretly pursuing to deliver packages by drone in 30 minutes. 

A pre-recorded demo showed an Amazon-branded “octocopter” carrying a small package off a conveyor belt and into the skies to a customer’s home, landing smoothly in the backyard, dropping off the item and then whizzing away. Bezos predicted a fleet of Amazon drones could take to the skies within five years and said, “it’s going to be a lot of fun.”

A decade later, Amazon is finally starting to launch drone deliveries in two small markets through a program called Prime Air. But just as it’s finally getting off the ground, the drone program is running squarely into a sputtering economy and CEO Andy Jassy’s widespread cost-cutting efforts.

…Jassy has resorted to trimming Amazon’s headcount, which grew massively during the Covid-19 pandemic, as he looks for ways to curtail expenses across the company. As part of his review, Jassy has zeroed in on some of Amazon’s more unproven bets, such as its Alexa, physical stores and robotics divisions. Now Prime Air is being added to the list of targets.

For Bezos, the staff reductions mark the latest setback in an ambitious project that’s been plagued with challenges.


The detailed timeline of how this has been hugely hyped and then quietly walked back is reminiscent of the whole metaverse thing. Speaking of which…
unique link to this extract

Microsoft has laid off entire teams behind Virtual, Mixed Reality, and HoloLens • Windows Central

Jez Corden:


Despite having acquired AltSpaceVR back in 2017, Microsoft culled the entire team behind the virtual reality workspace project this past week. As a result, AltSpaceVR will shutter for good in March, effectively ending Microsoft’s “metaverse” efforts with it. Supposedly, Microsoft Mesh will be AltSpaceVR’s successor, but it remains to be seen just how serious the company is about the so-called “metaverse,” despite CEO Satya Nadella’s buzzword-laden speeches on the topic at recent events. 

In addition to the death of AltSpaceVR, Microsoft has also culled the entire team behind the popular MRTK framework. MRTK (opens in new tab) is Microsoft’s “Mixed Reality Took Kit,” which is a cross-platform framework for spatial anchors in virtual reality spaces. MRTK was built for Unity VR integrations, and works with Meta’s headsets with a focus on HoloLens. 

HoloLens has been scaled back already in recent years following the departure of its chief architect Alex Kipman. Microsoft has been pursuing a HoloLens contract with the U.S. military, which was recently scaled back by the U.S. Congress, owing to reported problems with the program. 

For Microsoft to cull the entire team behind MRTK, which was due to release a new version just next month, it paints a picture of a company that perhaps no longer believes in virtual reality. There are many who believe the “metaverse” represents the next big opportunity in human-computer interfacing, but even Facebook, who rebranded its entire company to Meta in the belief of this technology, is scaling back in this area as well — laying off 11,000 staff back in November. 


How wrong I was to wonder last week about the future of the Surface line. That looks eminently safe. The metaverse/ARverse is still one of those things that looks technology-led, not user-need-led.
unique link to this extract

Deepfakes in the courts • Counsel, the law magazine

Lilian Edwards:


Three years ago, I suggested in a lecture for the Alan Turing Institute that although political deepfakes grab the headlines, deepfaked evidence in quite low-level legal cases – such as parking appeals, insurance claims or family tussles – might become a problem very quickly. We are now at the beginning of living in that future.

Much of what is written has concentrated on video evidence but audio may be a frontrunner growth area. In 2019, a lawyer acting in a UK custody dispute for the father resident in Dubai successfully challenged audio evidence apparently portraying the father as violent and aggressive. By gaining access to the audio files, forensic experts were able to point to the recording being a ‘deepfake’, which the mother had put together using online help fora.

Audio evidence is still relatively quick and easy to fake convincingly compared to video and photos and high-quality mass-market apps are already freely available to create ‘voice clones’. Videos variously claim that a voice can be cloned using only five minutes of recorded speech or even five seconds. Combine this tech with apps like Google Duplex designed to let you have an AI make calls for you, and audio deception becomes extremely simple. Audio deepfake detection is also advancing; however, it probably lags behind image detection (see Almutairi, Zaynab and Hebah Elgibreen, 2022. ‘A Review of Modern Audio Deepfake Detection Methods: Challenges and Future Directions’ Algorithms 15, no. 5: 155).

… in criminal cases, a provision under PACE [the Police and Criminal Evidence Act] requiring certification of computer evidence was abolished in 1999 and replaced by the common law presumption that a computer operates correctly in producing electronic evidence. This provision has become controversial, especially since the Post Office postmasters’ scandal, but its potentially important application to deepfake evidence has been little discussed. It is possible it sets up a presumption that video evidence is legitimate even where a deepfake is suspected.


We’re going to need independently sourced duplicates of everything.
unique link to this extract

“Gamification” • Soren’s Substack

Soren Iverson:


I saw someone tweet about a company who intentionally made their unsubscribe flow difficult in an attempt to retain customers. Unfortunately, dark patterns like this are not uncommon.

As I was thinking about this I started wondering, “what would happen if trying to unsubscribe became a game?”

Taking this to the extreme, what if there was a Minesweeper app where you had to beat Minesweeper to unsubscribe? Better yet, what if you added the constraint of only having one chance to beat the game each day?

By taking the nuance of dark patterns and pushing them to an absurd level, it highlights how wild the idea of designing friction into a process really is. It demonstrates the power of both good and bad design.


Scarily possible. The New York Times lets you sign up online, but of course you have to phone them up during their work hours, and they make it really hard. They’d probably like this idea.
unique link to this extract

‘Nightmare’: collapse of leading crypto lender traps investors • Financial Times

Nikou Asgari:


David [not his real name] is constantly lying to his mother. When she asks about the savings he is managing for her, he tells her not to worry. In reality, the $100,000 “nest egg” from the sale of her house is trapped at a crypto lending company.

“If I tell her, she’s going to have a heart attack,” says the 37-year-old from New York. “This was her everything.”

Eager to avoid rising inflation eroding his mother’s life savings, the television director last year placed the money with Gemini, the crypto exchange founded by the Winklevoss twins.

Gemini, run by Cameron and Tyler Winklevoss, offered a product called Earn that appeared to be an attractive haven for investors to leave their cash. Investors could earn more than 7% a year from the scheme when rates at traditional banks were close to zero.

David is now one of 340,000 Gemini Earn customers whose funds have been locked up after the group’s lending partner was wrongfooted by shockwaves that cascaded through the crypto market following the failure of Sam Bankman-Fried’s FTX exchange in November. Their plight has underscored the patchwork of often confusing regulations governing crypto in the US.

The Financial Times spoke to five users who said they believed it was similar to a savings account; in reality the product was a risky crypto lending strategy. “I thought I was just parking the money in a high-yield savings account and I can get it out anytime,” David said.

In exchange for the high interest rates, the Earn product lent out customers’ crypto coins. From February 2021, Gemini took retail investors’ funds and lent them out to crypto broker Genesis, which in turn loaned them to other digital asset market participants.

When FTX imploded, nervous investors rushed to pull their money from Genesis. The broker was unable to meet clients’ $827m worth of withdrawal requests, forcing it to suspend withdrawals from its lending business. On Friday, Genesis’ lending unit filed for bankruptcy.


At some point “David” is surely going to have to have a hard conversation with his mother, because that money isn’t coming back. Strange what a very large number of customers were sure that Gemini was as risk-free as a bank. Wonder how they came by that belief.
unique link to this extract

Britishvolt bankruptcy exposes Britain’s battery bluff • POLITICO

Joshua Posaner:


The EU — still a top destination for British auto exports — has agreed a zero-emissions mandate for cars and vans from 2035, and other major global markets are going the same way, starting with the US state of California.

While subsidies totalling more than €20 billion have been used to kickstart cell factory projects across the Continent, with similar amounts now flowing in the US under the Inflation Reduction Act green cash splurge, far less has been forthcoming in the UK.

The UK’s so-called Automotive Transformation Fund has earmarked just £1bn for all kinds of industry projects.

“These [cell] factories are being built in competitor countries,” Labour’s Reynolds said this week. “That is because they have governments with the vision and commitment to be the partner that private firms need to turn these factories from plans on paper into a reality.”

The UK government said it had committed “significant support” to Britishvolt in the form of a £100m investment that was contingent on management hitting certain milestones. That didn’t happen [the company has just filed for bankruptcy], so no public funds were spent on the project, the Department for Business, Energy & Industrial Strategy said.

The search is now on for another company to take over the site.

With battery-makers sniffing around for subsidies and qualified staff for new plants to serve surging demand from electric carmakers, juicier incentives from the UK government could attract interest.

For example, Sweden’s Northvolt is in talks with the German government about a plant near Hamburg, while Slovakia’s InoBat has plans for expansion. The world’s biggest battery player, China’s CATL, has global ambitions beyond its projects in Germany and Hungary. The UK could fit into that picture.

“The UK’s promise as an EV battery production location remains, with strong demand, a skilled workforce, and attractive manufacturing sites, all providing a compelling investment proposition,” said industry lobbyist Hawes.


You’d think serious questions would be asked about why Britain is so uniquely bad at this sort of thing. But people don’t want to engage with the problems.
unique link to this extract

Diffchecker: compare text online to find the difference between two text files


Diffchecker will compare text to find the difference between two text files.

Just paste your files and click Find Difference


One for the bookmarks. Programmers will be familiar with diffs, as they can show where things have been changed in large chunks of code.

The most recent use of this I saw was to compare the old version of Twitter’s developer agreement with its newer one, which indeed had been silently updated since its previous public version to ban third-party apps.

(Obviously, don’t paste files with sensitive content.)
unique link to this extract

Twitter is down to fewer than 550 full-time engineers • CNBC

Lora Kolodny:


Twitter’s full-time headcount has dwindled to approximately 1,300 employees, including fewer than 550 full-time engineers by title, according to internal records viewed by CNBC. Around 75 of the company’s 1,300 employees are on leave including about 40 engineers.

The company’s trust and safety team, which makes policy recommendations, design and product changes with the aim of keeping all of Twitter’s users safe, is down to fewer than 20 full-time employees.

…Before Musk led a $44bn leveraged buyout of Twitter last year, Twitter’s headcount stood at about 7,500 employees. Layoffs were rumored internally and expected to take place whether Musk’s takeover went through or not. However, Musk has cut Twitter personnel far more than many expected — or by about 80% according to the internal records and two recent employees who spoke with CNBC.

According to an engineer who resigned from the company, the loss of employees and reduced headcount will make it harder to maintain the service reliably while still building new features.


Ghost town in the office, increasingly a ghost town on the site. The loss of third-party app access and the user-hostility of the native app has driven people away. When Mastodon gets some proper apps, Twitter will have a problem. (Elon Musk responded to a tweet about this to say Twitter has about 2,300 employees, “hundreds” working on Trust & Safety and “several thousand” contractors.)
unique link to this extract

The lights have been on at a Massachusetts school for over a year because no one can turn them off • NBC News

Corky Siemaszko:


The lighting system was installed at Minnechaug Regional High School when it was built over a decade ago and was intended to save money and energy. But ever since the software that runs it failed on Aug. 24, 2021, the lights in the Springfield suburbs school have been on continuously, costing taxpayers a small fortune.

“We are very much aware this is costing taxpayers a significant amount of money,” Aaron Osborne, the assistant superintendent of finance at the Hampden-Wilbraham Regional School District, told NBC News. “And we have been doing everything we can to get this problem solved.”

Osborne said it’s difficult to say how much money it’s costing because during the pandemic and in its aftermath, energy costs have fluctuated wildly.

“I would say the net impact is in the thousands of dollars per month on average, but not in the tens of thousands,” Osborne said. That, in part, is because the high school uses highly efficient fluorescent and LED bulbs, he said. And, when possible, teachers have manually removed bulbs from fixtures in classrooms while staffers have shut off breakers not connected to the main system to douse some of the exterior lights.

…Minnechaug is the only high school in its district and serves 1,200 students from the towns of Wilbraham and Hampden. The original high school building, which dates back to 1959, was replaced with the current 248,000-square foot structure in 2012.

One of the cost-saving measures the school board insisted on was a “green lighting system” run on software installed by a company called 5th Light to control the lights in the building. The system was designed to save energy — and thus save money — by automatically adjusting the lights as needed.

But in August 2021, staffers at the school noticed that the lights were not dimming in the daytime and burning brightly through the night. “The lighting system went into default,” said Osborne. “And the default position for the lighting system is for the lights to be on.” Osborne said they immediately reached out to the original installer of the system only to discover that the company had changed hands several times since the high school was built.


unique link to this extract

• Why do social networks drive us a little mad?
• Why does angry content seem to dominate what we see?
• How much of a role do algorithms play in affecting what we see and do online?
• What can we do about it?
• Did Facebook have any inkling of what was coming in Myanmar in 2016?

Read Social Warming, my latest book, and find answers – and more.

Errata, corrigenda and ai no corrida: none notified

2 thoughts on “Start Up No.1935: Google calls in Page and Brin over ChatGPT, CNet pauses AI writers, Amazon cuts drone team, teeny Twitter, and more

  1. You can cancel NYTimes online now. Which is nice. But a little life hack for many subscriptions is paying via PayPal, and then just cancel the agreement at when you no longer wishes to pay for the sub.

  2. The move at Google reminds me that you have have the smartest people in the room, but paralysis sets in if you’re worried that it might impact your main revenue source (see also newspapers 1999 to 2010 with local ads). Hence probably why the funders are back: It’s not a technology problem Google has (their AI software is pretty remarkable), it’s a management problem.

    On a different topic, Microsoft surface laptops are quite well regarded by businesses, who are getting fed up with the quality of laptops from other vendors. You could argue they are the new IBM Thinkpad: Reliable, well built, and users love them. Hence it would be unlikely to be canned because they are effective at building loyalty to the brand (unlike the HoloLens products).

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.