Working from home has its own perils. Pets can be demanding, your back aches from hours at a desk, or you simply forget to move. There are a few apps that nudge you to move around or indicate that you’re not sitting in an ideal position, but they’re easy to dismiss.

I’ve spent the better part of a decade at a home desk, iterating on the setup as I go — gaming chair, lumbar support, the works. None of it guarantees good posture.

Then I came across Isa, a desk device from German startup Deep Care that takes a different approach entirely. It tracks posture, hydration, light, sound, and movement. And it does all of it without a camera or an internet connection, which, in an era of always-on surveillance, is a meaningful differentiator.

Here’s how it works and what’s inside. Isa has a 5.5-inch IPS HD screen and looks like a table clock. It is powered by USB-C; the company supplies a power unit with it, but you can use any of your existing chargers too, as it has a power consumption rating of roughly 2.45W.

The key sensor for the device is the Time-of-Flight (ToF) 3D depth sensor on the front — the same technology used in facial recognition and some smartphone cameras — that tracks posture and movement. It also enables beta features, such as counting the number of times you’ve had water or other liquids. The company said that the sensor works in the range of 0.15 meters to 1.8 meters. That means if the device is sitting on your desk, it can measure your movement, even when you stand up and move about. It also packs several other sensors: a ToF 1D sensor, a gyroscope, a barometer, a light sensor, a sound level sensor, a CO₂/VoC sensor, and a temperature and humidity sensor.

Getting started is straightforward — the device asks for a few details about you and your work routine. I found it strange that there was no option to set the device to India time (or any other Asian time zone). The company said Isa currently supports only EU and US time zones. Fair enough for now — but broader time zone support, or even a simple world clock, feels like a basic expectation for a desk device.

On the screen, Isa displays your posture with a squircle (a rounded square) ring that fills or empties based on how well you’re sitting, while a water-tank-style widget tracks your drinking. If you are not sitting in the correct posture, the indicator will turn yellow. The Apple Watch-style ring is a surprisingly effective nudge — when I see yellow or red, I straighten up almost instinctively.

The device vibrates to alert you if you’ve been slouching for too long, and I’m okay with that kind of mild shaming. That alert also indicates if you are leaning far too forward or back and helps you correct your stance.

A similar widget tracks movement, and if you have been stationary for a while, Isa suggests you get up, with on-device guided exercises to follow. When you return to your desk after a break, the movement tracker resets.

Deep Care chose not to include a cameras, which helps with privacy, but it comes with trade-offs.

If a bottle or some other object sits between you and the sensor, it may read that as a person and log you as stationary. Pets or housemates passing by can trigger the sensor, too. Isa usually figures out that you’ve stepped away and goes to a digital clock display, but I would have liked a manual button to tell it I’m not at the desk so it stops tracking.

Because of the sensor-only approach, the device occasionally told me I’d been stationary for too long when I’d been sitting for under half an hour. These are minor inconveniences. On balance, the device made me check my posture more often than I used to, and the exercise suggestions are truly useful.

To process all these features, the device uses a quad-core 2 GHz processor. The device can connect to Wi-Fi for software updates, but you can turn it off at any time.

Deep Care was founded by three former Bosch employees and initially sold Isa directly to businesses. It recently expanded to consumers — a shift that signals confidence in the retail market for workplace wellness hardware, and a test of whether a subscription model layered onto premium hardware can find a mainstream audience.

Isa is priced at €299 ($354) with two subscription tiers. The core plan (€4.99 per month) gives you access to posture tracking, healthy sitting habit tracking, drinking habit detection, and its exercise library. The Pro plan(€7.99 per month) lets you track light, noise, and CO2 levels for a healthy working environment.

The company plans to use Isa’s sensor suite to venture into mental health-related tracking. It claims that by using signals like posture, head movement, and chest movement, the device can measure breathing patterns. Plus, paired with environmental data like noise, light levels, and CO2 level, the company wants to introduce a stress-related score.

Even if you skip the mental health features, Isa is a solid device for anyone serious about posture and movement. It isn’t cheap, and the subscription adds to the long-term cost. But if you or someone you know works from home and has been meaning to do something about their desk habits, it’s one of the more thoughtful options out there.

OpenAI co-founder and president Greg Brockman is officially taking the reins of the company’s product strategy, according to Wired.

This seems to solidify an already-existing change, with Brockman overseeing OpenAI’s products on an interim basis while the company’s CEO of AGI deployment Fidji Simo is out on medical leave. Wired also reports that in a staff memo, Brockman described plans to combine ChatGPT and its programming product Codex into a single unified experience.

“We’re consolidating our product efforts to execute with maximum focus toward the agentic future, to win across both consumer and enterprise,” Brockman reportedly said.

This is just the latest OpenAI shakeup since CEO Sam Altman declared a “code red” at the end of last year and said the company needs to refocus on the core ChatGPT experience. Since then, OpenAI has halted “side quests” including video generator Sora and OpenAI for Science, and it’s been highlighting its ambitions to build an AI “super app.”

TechCrunch has reached out to OpenAI for comment. The company told Wired that Simo, who remains on medical leave, worked with Brockman on these changes.

Today, Cerebras Systems is a public company that sells AI chips for inference to giants like OpenAI and AWS. It held a blockbuster IPO on Thursday, with both of its co-founders billionaires, and ended the week worth about $60 billion.

But in 2019, when it was three years old, it came dangerously close to failure – incinerating a shocking amount of money. It was trying to solve a technical problem no one in the semiconductor industry thought could be done. 

“We were spending about $8 million a month,” founder CEO Andrew Feldman told TechCrunch of that period. “At this point, we had incinerated nearly $200 million trying to solve one technical problem.” 

Every few weeks, Feldman was forced to make the painful walk of shame to the board meeting to report another failure and more money burned. 

But he had no choice. Without a solution, Cerebras was dead anyway.

It was founded with an idea that was simple on paper. The microprocessor industry had spent its entire 50+ years making CPUs faster and cheaper by cramming more transistors onto a silicon wafer and dicing wafers into ever tinier pieces. But AI required so much compute power, many chips had to be strung together and then forced to communicate with each other. Cerebras’ founders believed turning a whole, even bigger wafer into one giant, powerful chip, would work faster. 

The problem was, no one had ever successfully done this before, for any reason, AI or not. Orchestrating that many microscopic electronic components onto a larger, but still thin, surface introduced compounding engineering problems. 

Once Cerebras crossed the first threshold of designing the mega chip and then manufacturing it with TSMC, the team hit the real roadblock. 

They couldn’t solve “packaging.” This involves everything after manufacturing the silicon itself: adhering it to a motherboard, getting power to it, dealing with heating and cooling as well as the pipes that would deliver and return data, Feldman said. 

Cerebras’ chips “were 58 times larger. We were using 40 times as much power as anybody had ever used,” he said. There were no premade heat sinks. No vendors. No manufacturing partners. The brightest minds in microprocessor engineering had tried for decades to build such big, yet more dense chips, and failed. 

The Cerebras team was left with trial and error in which “we destroyed an enormous number of chips” and an enormous amount of cash. But without functional packaging, the chip was useless. 

After exhaustive analysis of each failure, the team finally solved enough problems: how to cool it and move data around. In one instance, they had to invent their own machine that could bolt-in 40 screws simultaneously to secure the wafer to a board without cracking it. 

Feldman still remembers the day in July 2019 when it all, miraculously, worked.

They installed the packaged chip into a computer, turned it on and the entire founding team (pictured below) “just stood in the lab and stared at it,” he said. “Watching a computer run is about as exciting as watching paint dry. But there we were watching lights flashing on the computer, stunned that we’d solved this.” 

“That was one of the greatest moments of my life,” he said. That’s significant, because this same founding team had previously built and sold a pioneering cloud server startup, SeaMicro, to AMD for $334 million in 2012.

The day the chip finally worked was also about two years after OpenAI had talked to Cerebras acquiring it, which Feldman confirmed to TechCrunch occurred like the publicly revealed emails said it did. 

Those talks fell through amidst growing squabbling among the OpenAI founders, several of whom are angel investors in Cerebras. 

Today OpenAI is a customer and a partner, having loaned Cerebras $1 billion secured by warrants. Those warrants conditionally grant OpenAI about 33 million shares of Cerebras’ stock, the S-1 discloses. (33 million shares are worth over $9 billion at Friday’s closing price of $279.) 

Interestingly, Cerebras also agreed to not sell its wares to specific OpenAI competitors as part of that loan deal. Feldman wouldn’t confirm that the obvious company this involves: Anthropic. He did, however say that restriction is temporary. 

“It’s limited in time, and it was designed to make sure that we could get OpenAI the capacity,” he said.

The truth was, Cerebras hasn’t yet grown big enough to handle multiple fast-growing model makers anyway.  He likened selling AI compute capacity to an all-you-can eat buffet. Instead of trying to stuff itself on all potential customers, “We’re going to work with part of the buffet only, and we’re going to get comfortable with that, before we attack the rest,” he said.