OpenAI launched a new update to make its Chatbot less awkward.
OpenAI
OpenAI’s new GPT-Live model promises more natural human-to-AI conversations.
It will add “mmhm” or “sure” while the user talks — and nixes its signature pregnant pauses.
The update is part of a trend. AI labs are trying to make conversations feel less like passing batons.
OpenAI wants ChatGPT to stop waiting for its turn.
During a Wednesday livestream, the company unveiled GPT-Live, a new voice model powering ChatGPT Voice. OpenAI says the update is built to make talking with AI feel more like talking to a human.
Instead of waiting for a clean pause at the end of every sentence before responding, for example, the assistant can listen and speak simultaneously, OpenAI says. It will also acknowledge the user with an intermittent “mmhm,” “yeah,” and “got it” while the human is still talking.
In one demo, a user asked ChatGPT to fact-check the date of a coming meeting while also checking the weather and traffic along their route. ChatGPT responded with short acknowledgments like “hmm” and “sure,” then continued working through the request without losing the thread as the user added more information requests.
OpenAI also showed off ChatGPT’s ability to translate language in real time. Previous turn-based assistants had to wait for the speaker to finish before translating. Now, a full-duplex assistant — one that can listen and talk at the same time — can keep pace more naturally as a conversation unfolds, the company said.
Greg Brockman, OpenAI’s president, described the update during the livestream as a “much more natural way of interacting with your computer.”
OpenAI is not the only company trying to make AI voice assistants feel less like a baton-passing chatbot and more like an active listener. In May, Thinking Machines, the AI lab led by former OpenAI chief technology officer Mira Murati, teased similar technology. The company said its interaction models are designed to handle input and output continuously across audio, video, and text, instead of relying on the stop-and-start rhythm of traditional chatbots.
Thinking Machines said the updated models can “handle interaction natively” and smooth out human-to-AI interactions “rather than forcing humans to contort themselves to AI interfaces.”
The update also arrives as AI language models have become an internet punchline. Creators regularly mock their oddly chipper tone, their overuse of phrases like “awesome,” and the awkward little pause before an answer arrives.
NATO’s new strategy relies on uncrewed vehicles, like this Hunter Wolf unmanned ground vehicle, as a first response to an attack.
Spc. Mariam Diallo/US Army
New documents reveal NATO is building a vast AI network to deter Russian aggression.
It is part of a larger shift that foresees uncrewed systems as the first response to an attack.
The strategy “does not replace tanks, artillery, fighter aircraft, or soldiers,” an official said.
Along NATO’s eastern flank, the military alliance is no longer relying solely on tanks, fighter jets, and troops. Instead, it is building a vast digital battlespace network made up of thousands of sensors, drones, satellites, and artificial intelligence. The goal: to detect an attack on allied nations as early as possible and block the attacker before they can penetrate deeper into Alliance territory.
The concept is called the Eastern Flank Deterrence Initiative (EFDI). In documents obtained by BILD — which, like Business Insider, is part of the Axel Springer Global Reporters Network — one potential adversary is explicitly identified: Russia. It is intended not only to offset Russia’s advantage in mass and momentum — large troop numbers and rapid advances — but also to discourage Moscow from launching an attack in the first place. NATO refers to this principle as “deterrence by denial.” It is expected to rely on systems from Palantir and other leading Western defense contractors.
The new strategy creates a vast AI-guided network to detect and strike targets — not unlike the powerful web Russia has built fighting Ukraine. But it is just the beginning of a much larger shift, where uncrewed systems like attack drones and remote-controlled machine gun turrets will be NATO’s first line of defense.
The New Iron Curtain
Following Finland’s accession to NATO in 2023, the Alliance’s shared border with Russia expanded significantly. Today, NATO’s eastern border stretches from Finland through Estonia, Latvia, Lithuania, and Poland to Romania on the Black Sea. It also includes the border with Belarus, Russia’s close ally. Along this entire line, EFDI is intended to strengthen the Alliance’s defense through a system of sensors, drones, command networks, and conventional military forces.
Bild
For decades, NATO relied on the principle of “deterrence by punishment.” The idea was that if Russia attacked NATO territory, tanks, artillery, fighter aircraft, and ground troops would repel the assault and later recapture lost territory. These conventional capabilities will remain the backbone of NATO’s defense. Leopard 2 tanks, M1 Abrams tanks, HIMARS rocket launchers, artillery, and F-35 fighter jets will continue to play an essential role.
What is new, however, is an additional layer of defense. Its purpose is to detect, delay, and engage an adversary before NATO’s conventional ground forces come into contact with it, allowing those forces to preserve their combat power and be committed only at the decisive moment.
“EFDI does not replace tanks, artillery, fighter aircraft, or soldiers,” said Maj. Matt Blubaugh, a spokesman for US Army Europe and Africa. “It is designed to help preserve their combat power and give commanders more time and decision advantage.”
Bild
The New Strategy
NATO’s new strategy calls for uncrewed systems like drones to be the first response to an invasion or attack.
Staff Sgt. Reece Heck/US Army National Guard
When people think of border defense, they often imagine walls, anti-tank trenches, or soldiers guarding fences. EFDI is neither a wall nor a new front line. Instead, it is a distributed defense architecture supporting NATO’s entire eastern flank from Finland to Romania. The NATO documents frequently refer to a “Kill Web.” The term describes a tightly connected digital network. If one node fails, others immediately take over its functions.
Bild
These nodes include satellites in orbit, reconnaissance drones, radar systems, ground sensors, cameras, and electronic surveillance assets. Together, they continuously collect information about activity along NATO’s eastern flank.
In the recent past, engaging a newly detected target often took considerable time. A drone, for example, would first report the target to headquarters, where the information was analyzed before a firing order was passed through tactical commanders on to a military unit like a squadron. This process consumed valuable time.
In the near future, information collected by all NATO members is intended to flow into a shared digital network and be distributed immediately. Artificial intelligence will analyze the data in real time, helping commanders build a common operational picture as quickly as possible.
The new strategy is expected to incorporate technologies from a wide range of defense contractors through NATO’s open-architecture approach. At the core is Palantir’s Maven Smart System (MSS), which serves as the AI “brain” of the EFDI by processing data from all-domain sensors and enabling faster decision-making. The initiative also integrates systems such as Perennial Autonomy’s Merops AI drone interceptor, alongside capabilities from companies including RTX, Rheinmetall, Saab, Lockheed Martin and Boeing, all connected through the EFDI Data Backbone into a unified sensor-to-shooter network.
In practical terms, if a drone detects a Russian armored formation, the information will immediately be cross-checked with satellite imagery, radar data, and information from ground sensors. Commanders can then select which weapons — such as drones, artillery, rocket launchers, or other weapons — should engage the target. Weapons can be chosen by their range, their likelihood of hitting moving or static targets and also the importance of the targets.
NATO summarizes the principle in three simple steps: “See first. Decide first. Strike first.”
Machines Fight First
The front line itself is also expected to change significantly. Under NATO’s plans, uncrewed systems would be the first to confront an attacking force. A forward zone is envisioned where drones, ground robots, sensors, and other autonomous systems engage the enemy before conventional troops do.
The idea is straightforward: machines — not soldiers — should absorb the initial attack. This saves time and preserves NATO’s frontline combat formations.
“EFDI buys us time and clarity but at the end of the day, soldiers, tanks and aircraft are still needed to secure and hold ground,” Blubaugh said.
Bild
Lessons from Ukraine
The war in Ukraine has shaped the new concept. NATO is incorporating lessons learned by the Ukrainian military directly from the battlefield.
Specifically, thousands of low-cost drones, robotic systems, and sensors are intended to complement conventional weapons and offset Russia’s advantages in mass and momentum — its ability to field large numbers of troops and advance rapidly.
Luca-Marie Hoffmann is a BILD journalist based in Berlin, focusing on military affairs and crisis reporting.
The Axel Springer Global Reporters Network harnesses the resources of the company’s newsrooms to publish ambitious scoops, investigations, interviews, opinion pieces and analysis. It allows journalists — including those from POLITICO, Business Insider, WELT, BILD, Onet and Fakt — to collaborate on major stories for an international audience of hundreds of millions across platforms: online, print, TV and audio.
“The US built more new units in 2024 than any year in the past half-century, but that boom largely bypassed the Northeast and coastal California, which is exactly why rental competition there is so intense,” Zillow senior economist Kara Ng said.
Zillow ranked the hottest rental markets heading into the summer by looking at metros where rents increased quickly, vacancies dropped, and rental concessions — like waived fees or a month free off the rent — were rarely offered.
It shouldn’t come as a surprise that big cities like New York and San Francisco are regularly the busiest rental markets in the country; vacancies are low, and rents are climbing. However, Zillow found that one Northeastern city with a population that’s roughly 2% of New York City’s took the cake for hottest market: Providence, Rhode Island.
“In Zillow’s hottest rental markets, the math is simple: More people want to live there than there are homes to rent — whether for access to amenities, strong job markets or family ties, renters are competing over a limited supply,” Ng said.
Providence, Rhode Island, a city with a population of about 195,000, is small, and its size plays a factor in its ranking. It’s recently been labeled an unaffordable place to buy a home, mainly due to its lack of housing inventory, which, in turn, can drive up rental prices. Providence also has the lowest share of concessions in the top 10.
Read below to see the top 10 hottest markets for renters, according to Zillow.
The chain can be a really great place to find trendy decor, small furniture pieces, and a range of other household items that suit all sorts of styles.
As much as I enjoy shopping at the home-furnishing store, there are a few things I avoid getting there.
I get my larger furniture pieces, such as sofas, somewhere else.
Paynter Rhed
HomeGoods can be a great place to buy a stool or a small side table, but I wouldn’t buy larger furniture pieces here.
For example, many of the sofas and big chairs I’ve seen in the store look amazing from afar, but the quality hasn’t impressed me much.
Plus, at HomeGoods, you typically buy floor models. Your couch could be on the sales floor for weeks, with hundreds of strangers sitting on it or carts bumping into it at every turn.
There may be some hidden gems at a reasonable price, but when it comes to large pieces that will get lots of use, I’m not willing to gamble on quality. I’d rather shop at dedicated furniture stores, like a CB2 or West Elm, or seek out made-to-order options.
The candles are too hit-or-miss for me.
Paynter Rhed
The candle aisle at HomeGoods is my favorite part of the whole store. But as fun as the section is to peruse, the wide variety of brands, shapes, and scents can be a mixed bag.
I’ve found that many of the candles I’ve bought at the chain either don’t burn very well or lose their scent pretty quickly, even though they look great and smelled amazing in the store.
A lot of these candles are just fine in a pinch — and you might get lucky with a great one — but I’d rather spend more to buy something I’m confident is higher quality from an actual candle store.
Many of the large mirrors don’t pass my basic quality test.
Paynter Rhed
In the age of the mirror selfie, I completely understand the desire to add the perfect floor-length mirror to a space. After all, they’re essential for good outfit-of-the-day posts.
That said, the large mirrors I’ve seen at HomeGoods haven’t impressed me yet.
Mirrors can be massive statement pieces, so I suggest taking the time to look for high-quality options.
One way I like to test the quality of mirrors in a store is by standing far away from them to see whether my reflection gets distorted. If I start to look funky, I’ll pass on the mirror.
It can be hard to find a solid set of pots and pans at HomeGoods.
Paynter Rhed
Many of HomeGoods’ pots and pans look amazing — again, brands and quality vary — but I buy mine elsewhere.
Although you might get lucky and find name-brand items in the store’s aisles, I’ve bought many random pots and pans from here that looked nice but haven’t held up well.
Getting a full, matching set of cookware from HomeGoods is also tricky since not all stores carry the same pieces or multiples of a design.
If I’m supplying a kitchen from scratch or hoping to build a collection, I’d rather invest in a matching set from a trustworthy cookware brand.
Skip the kitschy letters and signs with quotes and phrases on them.
Paynter Rhed
Choosing the perfect finishing touches for a space can be one of the trickiest tasks. However, I’ll never waste money on hanging letters or signs with quotes. Quotes are meant for journals, not walls.
I’ve seen a lot of signs at HomeGoods with overused phrases and cringeworthy quotes. Some pieces with basic labels like “coffee” and “tea” even feel like they’re paying homage to farmhouse design in an untasteful way.
I suggest choosing text-free art pieces and decor items instead for a classier look. (Luckily, HomeGoods has a lot of aisles filled with these, too.)
Full-priced holiday decor is rarely worth it to me.
Paynter Rhed
HomeGoods has an impressive array of decor for every holiday and season I can think of — and it’s often on shelves long before whatever it’s celebrating begins.
However, as a frequent HomeGoods shopper, I’ve learned that if you wait until one day after the holiday, you can sometimes get up to 50% off on some seasonal items.
It can be risky, since the pieces you love may go out of stock before then, and sales aren’t guaranteed. However, the deals can be great and, as a planner, I actually prefer preparing for a holiday almost a year before it returns.
HomeGoods has some tasteful items for desk organization, but I still skip this section.
Paynter Rhed
HomeGoods is generally a great place to find one-off special items, but not so much coordinated or matching sets.
I like matching sets too much to consider buying desk decor or office supplies here, from letter organizers to pencil cups.
I’d shop elsewhere for finishing touches for my home office — maybe Target or even Ikea.
This story was originally published on December 21, 2023, and most recently updated on July 7, 2026.
Priyanka Devi Ramesh (left), Iren Azra Zou (center), and Udit Mehrotra (right) say many common assumptions about working in tech miss the mark.
Priyanka Devi Ramesh (left), Iren Azra Zou (center), and Udit Mehrotra (right)
Business Insider asked six tech workers about the biggest misconceptions in the industry.
Workers challenged assumptions about AI, Big Tech, and what their jobs actually involve.
Some said AI has changed the nature of tech work — but not in the ways many people think.
If you ask tech workers what people misunderstand about working in their industry, they’ll tell you — plenty.
In interviews with Business Insider, tech professionals from companies including Amazon, Google, and Snap challenged assumptions that their jobs are mostly coding, that AI has made their work easier, and that Big Tech is the only worthwhile career path. Some focused on misconceptions held by people outside the industry, while others pointed to misunderstandings from people already working in tech.
These professionals — including engineers, data scientists, and product managers — are working in an industry that’s changing rapidly. They say many people’s understanding of tech jobs hasn’t kept up.
Here’s what six tech workers believe are the biggest misconceptions about working in tech. (Their responses have been edited for length and clarity.)
My job is about much more than coding
Priyanka Devi Ramesh says working in tech involves much more than writing code.
Priyanka Devi Ramesh
Priyanka Devi Ramesh is a business intelligence engineer at Amazon. She’s 30 and lives in Virginia.
One of the biggest misconceptions is that working in tech is all about coding. People assume if you work in tech, you sit in front of a screen writing code all day. But my role as a business intelligence engineer is deeply rooted in understanding the business, talking to stakeholders, cleaning messy data, and telling stories through dashboards.
A huge part of my job is communication — translating complex data into something a non-technical customer or vendor can act on. Tech is far more cross-functional and people-oriented than most outsiders realize.
The perks are real. So is the pressure.
Sreeja Apparaju says working in tech is more demanding than many people realize.
Sreeja Apparaju
Sreeja Apparaju is a machine learning engineer at Snap. She’s in her 20s and lives in New York.
One misconception is that tech jobs are all hoodies, ping pong tables, and a four-hour workday. The perks are real, but they exist alongside the genuine intensity of on-call rotations, launch crunches, performance reviews, and the constant pressure to keep learning as the stack evolves under you.
There’s also the assumption that the work is purely solitary and screen-bound. The technical problems are only half of it; the other half is people: understanding what users actually need, negotiating priorities with partners, and communicating clearly enough that the right decisions get made even when you’re not in the room.
AI can’t do your thinking for you
Udit Mehrotra says AI can’t replace critical thinking.
Udit Mehrotra
Udit Mehrotra is a head of product at Amazon. He’s in his 30s and lives in Seattle.
The biggest misconception I come across right now is that you can outsource your thinking to AI. A lot of people have discovered that AI tools are genuinely impressive at producing output quickly, and the temptation is to let the tools drive.
The problem is that the quality of what comes out is almost entirely determined by the quality of thinking you put in. Garbage in, garbage out, except now it’s faster and looks more polished.
The bar keeps getting higher
Mike Kostersitz says AI hasn’t made his job easier.
Mike Kostersitz
Mike Kostersitz is a senior director, product management at Nike. He’s 60 and lives in Oregon.
A common misconception is that tech workers are coasting — making a lot of money for very little work because AI does the rest. That gets it backwards. AI doesn’t hand you free time; it removes the repetitive work that used to crowd out the important work.
The job hasn’t gotten easier — the bar has gotten higher. We’re expected to think more clearly, decide faster, and lead through more change in a quarter than we used to in a year.
Tech is much bigger than software engineering
Prerit Pathak says working in tech is about much more than software engineering.
Prerit Pathak
Prerit Pathak is a security engineer at Google. He’s in his 20s and lives in New York City.
I think many people mistakenly believe that being a “tech employee” is synonymous with being a software engineer. The reality is that technology is a vast ecosystem of roles like Product Management, UI/UX Design, Data Science, and Cybersecurity.
These specialists act as the architects, maintainers, and protectors of the digital world, ensuring that technical tools solve important problems for our future generations.
Big Tech isn’t the only path
Iren Azra Zou says Big Tech isn’t the only rewarding career path.
Rachel Wisnewski for BI
Iren Azra Zou is a software engineer at the trucking logistics startup Double Nickel. She’s in her 20s and lives in New Jersey.
I think some people in tech over-focus on working at the most famous tech companies. Those can be great, but they’re not the only path, and often not the best fit for everyone.
There are countless small and mid-sized companies doing interesting, meaningful work. I see a lot of people limit themselves because they think it’s either “Big Tech” or something boring and non-technical. That’s just not how the industry actually looks.
Do you have a story to share about how you’re navigating a career crossroads? If so, please reach out to the reporter via email at jzinkula@businessinsider.com, or via Signal at jzinkula.29.
Netflix has invested in getting YouTube shows like “Hot Ones: Extra Heat,” which features celebrities like Will Ferrell.
Netflix
Netflix is adding short shows from publishers like Condé Nast to further diversify its content mix.
Streamers have invested in snackable content to imitate YouTube.
This deal is “straight out of the YouTube handbook,” analyst Brandon Katz said.
Netflix is leaning into snack-sized content as it seeks to claw viewing time from YouTube.
The streaming giant is bringing more short videos to its app as part of licensing deals with major publishers, including BuzzFeed Studios, Condé Nast, Hearst Magazines, Penske Media, and People Inc., Netflix said Tuesday. Starting in early August, subscribers can watch videos ranging from 3 to 20 minutes from brands like Bon Appétit, Cosmopolitan, The Hollywood Reporter, Variety, and Vogue. The videos center on topics like travel, cooking, and fashion.
Netflix has been investing in video podcasts and short-form vertical video to better compete with YouTube, which has been rapidly growing its share of TV viewership in recent years. YouTube had a 13.4% share of US TV viewing in April 2026, versus Netflix’s 7.8%, per Nielsen.
Netflix’s new licensing deals are “straight out of the YouTube handbook,” said Brandon Katz, the insights director at entertainment research firm Greenlight Analytics.
“It’s another attempt to court low-cost engagement with lean-back pop culture programming,” Katz said. He added that “Netflix is trying to become a more habitual source of entertainment without having to pony up the premiums for hit-or-miss original programming.”
Other Hollywood streamers have been leaning into short-form clips over the last year, in hopes of driving higher engagement and training users to open their apps throughout the day, rather than just at night.
“These partnerships help us deepen fandom and create more ways for members to carry those stories with them throughout their day,” John Derderian, the VP of Animation Series and Kids & Family TV, said in a statement about the deal.
Short-form video isn’t new for Netflix. The streaming leader first experimented with vertical clips in 2021 with a “Fast Laughs” feature that showed snippets from comedy shows. Netflix shuttered the feed a few years later, before adding short-form video again earlier this year. The streamer has also had videos of less than four minutes as part of its “WWE Legend Profiles” series.
Netflix has been looking to speed up the addition of YouTube content to its platform by offering publishers more relaxed licensing terms, according to two people who’ve had direct conversations with the streamer in recent months.
During its initial push to add podcasts in 2025, Netflix insisted that podcasters pull their video shows off YouTube. That would be a tough ask for some creators and publishers who’ve built big businesses on YouTube. (It did some early deals with creators like Ms. Rachel that didn’t have that requirement.)
Now, Netflix is regularly talking about licensing shows and podcasts without requiring them to leave YouTube, the two people said.
“It’s a real change,” one person involved in the discussions said.
However, podcasters and other creators will still have to consider whether being available on Netflix will cannibalize their YouTube audience — and if the licensing fee Netflix pays will make up for it.