• Ukrainian commanders are having to train new combat troops at the front amid problems with training centers.
• There are advantages among the headaches with this kind of training, war experts say.
• And it's better than Russia's approach, which is pushing poorly trained troops into frontal assaults.

Ukrainian commanders are having to train new combat troops at the front due to training issues, but war experts say that even with the problems, it's better than what the Russians do.

While there's the possibility of a decrease in the overall quality of the Ukrainian forces as new combat troops replace experienced troops at the front, especially as newly mobilized forces come into play, the quality is still likely to be higher, at least for the time being, than that of the Russians, war analysts from the Washington, DC-based Institute for the Study of War said.

Ukrainian military officials recently told The Washington Post that newly recruited soldiers and those brought in from rear areas are learning how to fight on the front lines due to the lack of a "proper training system," according to a Sunday report.

Field commanders told the Post that because the training centers are inadequate, they have to spend several weeks teaching the basic fighting skills, including how to shoot.

Beyond the training center problems, the Ukrainian military is also, more generally, facing a dilemma: either train the new combat troops at the front or rotate out the experienced soldiers to train new ones.

Some partner nations have offered to help train the Ukrainian military, but this option requires sending Ukrainian troops to NATO countries. Doing so risks exacerbating existing delays, pulling commanders away, and limiting the number of soldiers Ukraine can place on the front lines.

With greater NATO support, though, NATO training could provide "further opportunities for Ukraine to improve its basic training mechanisms and improve the quality of newly deployed personnel." That situation, along with the other challenges, will not work itself out overnight.

Ukraine's training problems are likely to persist for some time, ISW analysts said in a Sunday assessment, further noting that "the average overall quality of Ukrainian forces on the frontline will likely decrease as experienced personnel rotate out and newly-deployed personnel reach the" front lines.

The analysts noted, however, that "new soldiers will likely learn rapidly as they fight alongside experienced veterans."

The "decisions to train newly-deployed personnel on the front before committing them to combat indicates that the overall quality of Ukrainian forces will likely remain higher than that of Russian forces in the near- to mid-term," they said, as Russia has problems of its own.

Russia has been building up its infantry quickly with a range of different soldiers from penal convicts to volunteer soldiers. Because Russia has prioritized quantity over quality, the rapid deployment of these soldiers with minimal training has resulted in significant casualties.

"The Russian force generation mechanism has largely met the replacement rate of casualties in Ukraine," ISW said, but it prioritizes "fast redeployments of new personnel for additional 'meat' assaults over effective training."

ISW analysts also pointed to reports indicating that Russia is suffering from a lack of training instructors, which has resulted in continuous training delays.

"Russian milbloggers have consistently complained about ineffective Russian training since partial mobilization in September 2022," ISW reported. "A former Russian Storm-Z instructor recently claimed that Russian 'strategic' reserves are 'doing nothing for months.'"

• I booked the cheapest interior cabin for my four-night cruise on Carnival's new Carnival Firenze.
• It was ugly, dated, and lacked decor, but shockingly spacious. 
• Interior staterooms in Carnival's newest ship start at $90 per person per day in 2024.

One of the best parts about cruising is feeling the open ocean's cool breeze rush around your skin.

So imagine how grumpy I was when I had to book a windowless interior cabin for my most recent sailing — on my first Carnival cruise, no less.

As a travel reporter, I go on several of these vacations at sea every year. I'm usually put up in balcony staterooms, the most popular cabin category.

Yes, I've been spoiled. And yes, that means I've always feared inside cabins.

Cruises can be overstimulating and overwhelming. Private balconies are the only place I've gotten peace, quiet, and fresh air on some ships.

So, unsurprisingly, I was dreading the interior cabin I had booked for my four-night Carnival Firenze cruise in mid-May.

No fresh air? No cool ocean breeze? The horror!

I expected a small, dark, moldy dungeon deprived of life, happiness, and light.

Turns out, I was being a bit too dramatic. Who would've guessed?

I'll admit was pleasantly surprised.

But before I get to the positives (spoiler alert: it's size), I have to address the negatives.

It might've been the ugliest cabin I've ever stayed in. And that's coming from someone who's sailed on a Margaritaville cruise.

To understand the stateroom, you must understand the ship's relatively short history.

Firenze joined Italy-based Costa Cruises' fleet in 2020. The company, owned by Carnival Corp, had planned to launch the ship in China but couldn't because of the COVID-19 pandemic.

Firenze's stint with Costa was then shortlived. In 2022, Carnival announced it would acquire Firenze and its sister ship, Venezia, to grow its US footprint amid "strong interest in people wanting to sail with us," a Carnival spokesperson told Business Insider.

The spokesperson said Firenze underwent a two-month refresh to 'install the Carnival Fun Italian Style Concept.'

That would explain the otherwise unnecessary photo of Florence, Italy behind my bed.

The print was both the only decor and the only reference to Italy in my dingy cabin. Unfortunately, it looked both tacky and misplaced.

It did, however, disguise the ominous "pull the cord for assistance" feature.

It's 2024. Why weren't there any bedside outlets?

The ship is less than four years old. The cabin looked quadruple its age, no thanks to the dated furniture that all came pushed against the wall.

The chaise's greyish pattern clashed with the yellow and red eyesore of a rug. And the small wood and faux-marble side table looked cheap and misplaced.

Gripes about aesthetics aside, everything in the joint living room and bedroom was perfectly functional.

On the bright side, the mattress was comfortable. And the large desk and closet could have provided enough storage for more than two travelers.

The room also had motion-sensing lights mounted by the floor, a necessary amenity for a space devoid of natural light.

Plus, the unappealing decor encouraged me to spend more time around the ship's common spaces.

But if you think the bedroom looks old, wait until you see the bathroom.

It was about as glamorous as a gas station restroom.

Most of my cruise cabin bathrooms have had modern, cool-toned decor, like the one pictured on the left. Carnival seems to have avoided the industry trend as much as possible.

The yellow and red tiles were giving vintage McDonald's.

Unbeknownst to me, the cruise line had put me in one of the ship's accessible, wheelchair-friendly accommodations. This meant the shower (as in, a curtain and a few drains on the floor) was one of the most spacious I've had at sea.

However, it only came with a two-in-one body wash and shampoo. Apparently, both Royal Caribbean and Carnival seem to have a BYOHC (bring your own hair conditioner) policy.

Some storage would've been nice, too.

An unnecessary number of stacked tissue boxes and toilet paper rolls occupied the only shelves in the bathroom.

There was no need for these excess paper products. My cabin attendant already cleaned and restocked my stateroom once a day, as is typical with Carnival.

I'm a girlie with an extensive skincare routine and a compulsive need to organize my products. With almost no storage in the bathroom, I needed these shelves.

With all these complaints, you may wonder why I said I was pleasantly surprised by my stateroom.

It's all because of one major feature: Its size. I never felt claustrophobic despite being stuck in a windowless 150-square-foot box.

The bathroom, one of the most spacious I've seen on a cruise ship, was so big that it could comfortably fit a family of four.

Back in the living room, I could've rolled out a yoga mat and worked out without moving anything around.

Heck, I could've comfortably starfished on the floor without hitting any furniture. I don't know if I can say the same for most of my previous cruise accommodations.

Would I have preferred a balcony or window? Yes.

Did it feel more like a hospital room than a floating hotel room? Yes.

However, for the spaciousness, the cost is hard to beat. As a solo traveler, I paid $735 for my cabin, including $64 in optional gratuities.

In 2024, the cheapest interior cabin for two guests on Carnival Firenze starts at about $360 per person, including taxes and fees, for the same four-night itinerary as the one I went on.

That's $90 a night for a cabin, food, onboard amenities, and stops at Catalina Island and Ensenada, Mexico.

Plus, it's much cheaper than the newest ships of other cruise lines (I'm looking at you, Royal Caribbean).

So, if you are desperate for a dirty and cheap getaway, and Carnival Firenze has been catching your eye, don't be afraid of the interior cabin.

Yes, it isn't lovely. But at least it's completely functional and large.

If I — spoiled — can do it, so can you.

• Jim Ward, 61, said he "escaped" to Tennessee after living in Texas for 20 years.
• Ward said he was ultimately tired of Texas' hot climate and relatively higher property taxes.
• Ward had to pay more for his Tennessee house but loves having easy access to hikes in the mountains.

This as-told-to essay is based on a conversation with Jim Ward, a 61-year-old who works in cybersecurity and moved from Dallas, Texas, to Knoxville, Tennessee, in 2022. The essay, which also incorporates quotes from emails between Ward and BI, has been edited for length and clarity.

I am not a native Texan, but I have lived there during three different time frames of my life: Houston from 1991 to 1993, Dallas from 1994 to 1995, and then Dallas again from 2002 to 2022.

Texas has never really factored into any of my life plans, but job and family situations just kept pulling me there.

When I relocated to Texas in 2002, I intended to stay for only four to five years so that I could attend a graduate school program in Dallas and then move elsewhere. But my oldest had just started middle school, and had become very active in school and our church. I did not want to uproot my kids, so we stayed.

After 20 years of living in the Dallas-Fort Worth area, my wife and I finally escaped to Tennessee two years ago.

Texas has its pros and cons

Over the years, I have developed a love-hate relationship with Texas.

There are definitely a lot of positives about living in the state. For instance, it has many great public schools and restaurants, especially Tex-Mex ones. And from a financial standpoint, its biggest advantage is that there is no state income tax — though that is more than offset by property taxes.

There are problems though, like oppressive heat and humidity, a lack of natural beauty, and enjoyable outdoor activities.

There's also a lot of distance between key destinations, causing far drives.

It's way too hot in Texas and there's not much to do

One of my primary problems with Texas is its climate and weather.

It gets quite hot during the summer, and throughout the year there are also severe ice and hailstorms. Sometimes it can result in periods of extended power outages during both summer and winter months.

As far as geography goes, in the Dallas area, the terrain is predominantly flat with few trees. Consequently, outdoor activities are limited, and there are few opportunities for hiking and camping.

The lack of foliage also means there is minimal shade. Even activities like running are affected by the relentless heat. I always had to go early in the morning or later in the evening.

That said, there are many lakes, and people do like to go boating or water skiing.

Texas has a great tech scene

Just about every major corporation in the US has a presence in the Dallas-Fort Worth area — all of them with their own internal technology needs.

As a result, there are numerous technology companies, particularly in cities like Las Colinas and Plano. This makes DFW a great place for tech workers because there are many professional networking opportunities, such as associations, events, and conferences.

While living in Texas, I had a job that required a lot of traveling. DFW's central location was a plus because it was convenient for air travel.

The area has two major commercial passenger airports: Dallas Fort Worth International Airport and Dallas Love Field Airport.

Depending on your destination, numerous air carriers offer direct nonstop flights to almost anywhere, greatly reducing travel time — though it may limit your frequent-flyer miles.

Taxes are higher in Texas but the homes are more affordable

We owned and lived in four different houses across Texas, while also investing in two houses in Irving and Fort Worth, which we rehabilitated and rented out. We sold both of those rental houses shortly before we left Texas.

The last home we owned and lived in was in Highland Village, a suburb of Dallas-Fort Worth, which was our favorite. My wife and I purchased our 3,600-square-foot home in 2015, and it appreciated by 50% over six years.

Our property taxes only went up by about $1,000 since we purchased our home, but were still significantly higher than in the previous states I lived in, Virginia and North Carolina.

The real estate market favored us when selling the Highland Village house, but we weren't as lucky when purchasing a new house in Tennessee. We ended up buying a home that was 1,000 square feet smaller but 30% more expensive than the house we sold in Texas.

We chose Tennessee because our two daughters moved here. We live in a suburban area fairly close to downtown Knoxville. We're in an apartment as we wait to move into our new house.

Our quality of life has improved in Tennessee

There are certainly several similarities between Tennessee and Texas, but I also see a lot of differences.

For instance, while the summers can be quite hot here, the weather doesn't stay that way for long. Tennessee actually experiences all four seasons.

In Knoxville, there are a lot of the same amenities that you might find in a city like Dallas, but there are also a lot more outdoor activities.

My wife and I have really enjoyed hiking and being near the Great Smoky Mountains National Park — the trailhead is about 50 miles from our home in Knoxville. We visit the national park numerous times throughout the year.

Knoxville is a sizable city, but it's nowhere near the size of Dallas or Houston, so that's been a big change. Even still, we've enjoyed living here so far.

I never really enjoyed downtown Houston or Dallas. But in Knoxville, we go downtown almost every weekend — there's a lot you can do without leaving the city.

It is very "walkable" and easily accessible, with numerous dining and entertainment options as well as frequent special events. We have season tickets for Broadway musicals at the Tennessee Theatre.

While there are some things about Texas I miss, like the abundance of air travel options and its restaurant scene, I do wish we had moved to Tennessee a little earlier.

• UC Berkeley sees a 48% jump in first-year applications to study computer science.
• Despite generative AI advances, students are eager to pursue computer science careers.
• Human developers remain essential for creating something new.

One of the most persistent concerns around generative AI is whether the technology will put workers out of a job. This idea has particularly caught on in the context of software coding.

Github Copilot can write a lot of code these days, so is it even worth studying computer science now? That's been a question on the minds of math-minded high schoolers since ChatGPT burst on the scene in 2022.

There's a new data point that helps answer at least part of this question: Students are still lining up in droves to take computer science in college.

An eye-popping data point

Let's take The University of California Berkeley as an example, as this college at or near the top for computer science.

First-year applications to UC Berkeley's College of Computing, Data Science, and Society CDSS increased 48% this year. There were 14,302 (non-transfer) applications for these CDSS majors in the Fall 2024 incoming class, versus 9,649 the previous year.

For context, the number of first-year applications to UC Berkeley as a whole didn't change much from a year earlier.

This was announced last week by Professor Jennifer Chayes, the dean of Berkeley's College of CDSS. She popped these eye-popping stats during a fireside chat with Governor Gavin Newsom and Stanford Professor Fei-Fei Li at the at the Joint California Summit on Generative AI in San Francisco.

There's a role for human software developers

Afterwards, I got in touch with John DeNero, Computer Science Teaching Professor at UC Berkeley, to talk about this some more.

He's also chief scientist at Lilt, a generative AI startup, and he was previously a researcher at Google working on Google Translate, one of the first successful AI-powered consumer apps.

"Students express some concern that generative AI will affect the software engineering job market, especially for entry-level positions, but they are still excited about careers in computing," he wrote in an email to Business Insider. "I tell them that I think many of the challenging aspects of software development can't be performed reliably by generative AI at this point, and that I expect there will still be a central role for human software developers long into the future."

AI can't do new things very well

Generative AI is currently very good at replicating parts of software programs that have been written many times before, DeNero explained.

That includes computer science homework assignments! See BI's coverage on how much ChatGPT is used to cheat on homework.

What if you want to create something new? This is where smart human coders will still be needed. (This makes logical sense as AI models are trained on data. If that information doesn't exist yet or it's not part of the training dataset, the models often get in trouble).

Generative AI "requires a lot of thoughtful human intervention to produce something new, and all consequential software development projects involve quite a bit of novelty," DeNero said. "That's the hard and interesting part of computing that currently requires clever and well-trained people."

"Generative AI can speed up the more mundane parts of software development, and software developers tend to adopt efficiency tools quickly," he added.

What happens at Lilt?

This applies to what's happening at Lilt, which is building an AI platform for translators.

Google Translate first came out 18 years ago. And still, human linguists have jobs and are relied upon when translations are really important. For instance, you can use Google Translate to read a Japanese train timetable maybe, but would you use the app to translate your business's most important contract without having a human expert check it? Probably not.

"To reliably produce publication-quality translations, human expert linguists are still at the center of the process, but by using Lilt's task-specific generative AI models, those experts are much faster, more accurate, and more consistent," DeNero said. "As a result, more text gets translated at higher quality into more languages."

He expects this same pattern to play out in software development: A small team of highly trained human developers will have an even greater capacity to build useful high-quality software.

"And so, future Berkeley graduates will have plenty of opportunities to use their computing skills to improve the world," DeNero said. "Hopefully some more of them will come work for Lilt."

• Israel is considering mounting rotary cannons on armored vehicles to shoot down drones.
• Israel may put the Vulcan cannon used on F-16s to an armored infantry carrier.
• Ukraine uses autocannons and heavy machine guns to target incoming drones.

Ukraine and Israel are exploring cost-effective ways of countering drones as they both face repeated attacks from their determined adversaries.

Israel is exploring the feasibility of mounting the six-barrel M61 Vulcan cannons found on American fighter jets atop armored vehicles deployed in the country's northern border regions with Lebanon. Iran-backed Hezbollah has tripled the number of drone strikes against north Israel in the past three months, forcing Israel to seek an effective counter without resorting to missiles.

Israel has one of the world's most advanced air defense networks but is still having trouble countering these drones. The M61 may prove a viable, cost-effective solution for point defense of armored columns and ground troops.

Ukraine is at the forefront of drone defenses. Its troops have devised a novel solution for countering Russia's Iran-supplied Shahed drones. Kyiv is using a vast network of sensors to detect Russia's low-flying drones and passing targeting data to heavy machine guns and flak cannons mounted on vehicles throughout the country that then attempt to shoot them down.

There are similarities between Ukraine's strategy and Israel's new approach, said Federico Borsari, a defense expert at the Center for European Policy Analysis (CEPA).

"In principle, yes, this effort resembles what Ukraine has done with the creation of mobile anti-drone teams with heavy machine guns, powerful searchlights, and other sensors to counter Shaheds and other low-altitude slow-flying drones," Borsari told Business Insider.

"This type of capability — which is relatively simple in nature — will be relevant for the counter-unmanned aerial system (C-UAS) task in the future but is not the only one, and it will be integrated into a broader set of other capabilities," Borsari said.

Ukraine needs a cost-effective solution to defeat the thousands of Russian one-way drone attacks that have forced Kyiv to expend large quantities of its expensive air defense missiles to shoot down. German-supplied Gepards, a cannon design dating back to the 1960s, have proven highly effective in countering Russia's drones and are a cheaper solution than more modern air defenses like the NASAMS; each NASAMS missile costs about $1 million. However, Ukraine faces severe difficulties acquiring more ammunition for the Gepard automatic cannons.

"I think the Israelis are likely seeing what does and does not work in Ukraine — the key is to create an effective and relatively cheap defense against incoming drones, and Ukrainians are one of the leading developers in that area right now," Samuel Bendett, an expert in unmanned and robotic military systems at the Center for Naval Analyses, told Business Insider.

"Another aspect to consider is that such weapons may work against specific drone types but not others," Bendett said.

CEPA's Borsari believes the M61 cannon could be an "interesting solution" for countering certain drones.

"In principle, its use in both fixed and mobile applications is not new and is conceptually similar to many other anti-aircraft guns that have been used for counter-drone purposes in Ukraine," Borsari said.

"The integration with radar or electro-optical/infrared (EO/IR) tracking sensors can certainly improve its accuracy, but this weapon is still intended for close-range C-UAS missions," Borsari added. "The radar-guided Phalanx close-in weapons system, for example, has been in use for decades by the navies of the US and other countries."

The M61 is carried by jet fighters like the F-16 Fighting Falcon and the F/A-18 Hornet that can fire the 20mm rounds at targets up to nearly two miles away. It has the power and rate of fire to destroy incoming drones, but its cost-effectiveness turns on how many rounds it must shoot to do so.

"The longer the burst, the higher the cost per interception due to the use of more rounds," Borsari said, noting it costs $180,000 to fire the M61 for one minute, during which time it can fire 6,000 rounds.

"Furthermore, engaging faster drones could be difficult," Borsari said. "There are now short-range ground-launched missiles and guided rockets meant for C-UAS that cost around $40,000 each, along with reusable ones equipped with a high-power microwave payload (Coyote Block-3) to defeat swarms at short range."

"So, in general, the M61 can be a solution, but it's not the only one."

Israel previously integrated M168 Vulcan cannons and FIM-92 Stinger missiles on its Machbet short-range air defense system, which it developed in the 1990s based on the American M113 armored personnel carrier to defend infantry from air attack. Its present evaluation of the M61 may be the first step to developing something similar to deal with low-flying drones.

Interestingly, an unmanned Israeli M113 was reportedly sighted in southern Gaza in recent days.

Borsari believes the M113 is one type of vehicle that can be "repurposed" as an unmanned ground vehicle (UGV) to defeat drones.

"Soon, UGVs will certainly be used in the C-UAS configuration and be progressively integrated into maneuver formations, although they won't necessarily be the standard mobile air defense capability," Borsari said.

"Rather, they will be part of a diverse and layered set of capabilities for C-UAS attached to ground formations, some crewed, some uncrewed, and equipped with different effectors — both kinetic and non-kinetic — including heavy automatic guns, short-range missiles, directed energy like high-power microwave, and electronic warfare."

• Jeremy Cullifer sold part of his backyard through startup BuildCasa for $100,000.
• The grocery wholesaler, 46, plans to use the money to upgrade his 70-year-old house.
• Cullifer said California's increasingly bad housing crisis was a big motive for splitting his lot.

This as-told-to essay is based on a conversation with Jeremy Cullifer, 46, about his decision to sell a portion of his backyard through startup BuildCasa to both make money and contribute to solutions to California's affordable housing crisis. The conversation was edited for length and clarity.

I've been in Sacramento my whole life.

I've been in my house for 19 years. I bought it through the family for $180,000 after my great-grandmother passed. She bought it in 1953.

It's a three-bedroom, one-bath house. It's 1,096 square feet on a quarter-acre lot. I have a moderate front yard that wraps around because it's a corner lot, and then a large backyard.

BuildCasa had put out a mailer to my mother — she lives in the same community on a different street.

She said, "You should give some thought to this. You don't do anything with your backyard and you're always complaining about affordable housing for people in Sacramento. You might as well be part of the solution instead of part of the problem."

I'm the only person I know in my peer group that owns a home. Everybody else is renting.

My son is about to get his first apartment, and it's $1,500 a month. It's insane. The supply of available housing is nowhere near the demand, so of course the prices just keep going up and up.

My daughter moved to New York City after she graduated college and struggles with affordable rent. That really was the biggest eye-opener to me.

I sold my backyard for $100,000

A couple of years ago, California Senate Bill 9, or SB 9 — which allows single-dwelling units to subdivide their property — was passed. My understanding is that the subdivision has to be at least 40% of the existing land in order to be eligible.

What BuildCasa is going to do is create a new property line for me, which is about 10 feet off my back porch, and then the remainder of the backyard would then become its own plot of land. It would become its own separate lot with its own separate address, and a developer would build a stacked duplex on it.

I was intrigued at the very beginning, and then I had a meeting with BuildCasa and did some research online. Then I talked to my mom, my brother, and an aunt and uncle — because it's been a family home since the '50s, so I didn't want to create family strife by changing the homestead.

They were indifferent. They were like, "What are you doing to the house?" Well, I'm not doing anything to the house. I'm just going to have less backyard.

BuildCasa laid out a whole bunch of options and let me decide what was best. They checked in with me five or six times to make sure it was actually something that I wanted to do — it was not just a knee-jerk reaction.

They laid out what was possible and a bunch of different paths. I picked what worked best for me, what I was most comfortable with. We ended up listing the property and letting people bid on it. I had a couple offers that I had rejected, and then we ended up partnering with a developer.

I sold it for $100,000, of which I expect to pocket $30,000. I'll have to pay around $35,000 to my mortgage holder — because of the loan-to-value amount is going to change, I'm probably going to have to pay down to my mortgage company to get that back-end ratio.

And then the remainder will be paying BuildCasa for all their time, effort, energy, knowledge, expertise, and going through the process.

I'm selling to the developer, and BuildCasa is facilitating both that relationship as well as going through all the red tape and the government requirements necessary to get the permits, go through the zoning changes, and all that kind of fun stuff.

Once you've divided and sold your property, there's no going back — unless you buy it out from somebody else, and that would be a financial nightmare.

The extra cash is nice, but I think more people will participate once the housing crisis gets worse

There were a number of reasons I did this.

Number one: maintenance and upkeep. I hate yard work, and now I'm going to have to do a lot less of it. That's pretty exciting.

Anybody who tells you they're just solely altruistic is lying to you.

Number two: the money I get from that sale, I'll be able to upgrade my house.

It's a 70-year-old home. It could use a new bathroom, it could use a new kitchen, and I'll be able to do that without financing it, which is pretty exciting. It'll improve my own personal living situation while still contributing to being a part of the solution for housing.

I talked to a few community members about what I'm doing. I've been here for a large portion of my life, so I know people pretty well. I didn't really have anybody who had any significant concerns I talked to.

I think the trend that has grown over the last 20 years of people buying property for investments and flipping them and continuing to raise the prices has really made it almost impossible for anybody less than the upper-middle class to buy a home.

At some point, there'll be a tipping point when people's kids can't buy a house or their nieces and nephews can't buy houses, or their best friend can't buy a home, and people will start to wake up to the scope of the problem.

We've had quite a few ADU-type buildings in the neighborhood, but I have not seen anybody do a subdivided lot in this community yet.

I would be shocked if it didn't catch on.

• Is YouTube's massive content library being used to train its AI models?
• CEO Neal Mohan says some creators' contracts with the platform mean their content could be used.
• He also said that if OpenAI trained Sora on the video within YouTube content, it would violate the terms of service.

Google is betting big on AI. AI models, including Google's Gemini, require a ton of training data to be competitive.

So a natural next question is: Is Google using its massive trove of YouTube videos to further those AI ambitions?

To help try to answer that question, we looked at what YouTube's CEO has said on the topic and what the platform's terms of service say, along with sending some clarifying questions to its parent company Google.

YouTube CEO Neal Mohan was asked in an interview about the possibility of Google using YouTube's massive digital content library to train its AI models.

In April, The New York Times reported that "Like OpenAI, Google transcribed YouTube videos to harvest text for its A.I. models, five people with knowledge of the company's practices said. That potentially violated the copyrights to the videos, which belong to their creators."

Mohan said some YouTube creators have specific contracts that can allow their content to be used in AI training.

"Google uses YouTube content really in accordance again back with those terms of service or individual contracts that we might have with creators or uploaders to our platform," Mohan told Bloomberg's Emily Chang in a full-length interview, some of which was first published in April.

"Lots of creators have different sort of licensing contracts in terms of their content on our platform, lots of rightsholders do," he added.

Basically, it sounds like YouTube's CEO is saying that any AI training the company is doing with YouTube content, whether it's scraping video titles or transcripts or the video content itself, is being done in a fashion that honors the terms that content creators have agreed to.

"And so some portion of that YouTube corpus may be being used for those models, but it's going to be in concert with whatever the terms of service or the contract that that creator has signed before uploading their content to YouTube," Mohan said.

As The New York Times reported, Google may not be the only company looking to YouTube for AI training data.

In a March interview with The Wall Street Journal, OpenAI CTO Mira Murati was asked if the company's AI text-to-video generation tool Sora was trained on YouTube content. She responded that she was "actually not sure about that."

Mohan says that, depending on what, if any, data OpenAI scraped, it could violate YouTube's terms of service.

"Our terms of service does allow for some YouTube content like the title of a video or the channel name or the creator's name to be scraped because that's how you enable the open web for that content to show up and maybe show up in other search engines or what have you and be available that way," Mohan told Chang.

"But it does not allow for things like transcripts or video bits to be downloaded and that is a clear violation of our TOS," he said.

YouTube's terms of service define content as "videos, audio (for example music and other sounds), graphics, photos, text (such as comments and scripts), branding (including trade names, trademarks, service marks, or logos), interactive features, software, metrics, and other materials whether provided by you, YouTube or a third-party."

The terms say uploaders "retain ownership rights in your Content" but also "grant certain rights to YouTube and other users of the Service."

"By providing Content to the Service, you grant to YouTube a worldwide, non-exclusive, royalty-free, sublicensable and transferable license to use that Content (including to reproduce, distribute, prepare derivative works, display and perform it) in connection with the Service and YouTube's (and its successors' and Affiliates') business, including for the purpose of promoting and redistributing part or all of the Service," the terms say.

So while Mohan's remarks and YouTube's terms of service shed more light on the matter, it's still not entirely clear if and how the average YouTube video could be used by Google for AI training purposes.

Business Insider reached out to Google to ask if the company was training any of its AI models or products, such as its recently announced text-to-video generator, Veo, on the actual video files of YouTube content. We'll update this post if we hear back.