OpenAI transformed its popular ChatGPT service into a powerful search engine today, marking the company’s boldest move yet to compete with Google. The upgrade lets users ask questions in plain English and get real-time information about news, sports, stocks, and weather — features that until now required a separate search engine.

“We believe finding answers should be as natural as having a conversation,” an OpenAI spokesperson told VentureBeat. The company will roll out the feature first to paying subscribers, with plans to expand to free users in coming months.

ChatGPT Search: How OpenAI’s new AI-powered web search actually works

Unlike traditional search engines (i.e. Google and Bing) that return a list of links, ChatGPT now processes questions in natural language and delivers curated answers with clear source attribution. Users can click through to original sources or ask follow-up questions to dig deeper into topics.

The technology builds on OpenAI’s SearchGPT experiment from July, which tested the search features with 10,000 users. That limited release helped the company refine how its AI processes web information and attributes sources.

The system runs on a specialized version of GPT-4o, OpenAI’s most advanced AI model. The company trained it on massive amounts of web data and fine-tuned it to understand context across longer conversations.

Major news publishers partner with OpenAI to power next-generation search results

Major news organizations including the Associated Press, Axel Springer, and Vox Media have partnered with OpenAI to provide content. The deals aim to address long-standing concerns about AI systems using publishers’ work without permission or payment.

“ChatGPT search promises to better highlight and attribute information from trustworthy news sources, benefiting audiences while expanding the reach of publishers like ourselves who produce premium journalism,” said Pam Wasserstein, President of Vox Media, in a statement. Publishers can opt out of having their content used for AI training while still appearing in search results.

Inside OpenAI’s $5 billion bet on custom chips and AI infrastructure

The launch comes as OpenAI races to build its own technology infrastructure. The company recently announced deals with AMD, Broadcom, and TSMC to develop custom AI chips by 2026 — a move to reduce its reliance on Nvidia’s expensive processors.

These investments don’t come cheap. Microsoft, OpenAI’s biggest backer with nearly $14 billion invested, said this week the partnership will cut into its quarterly profits by $1.5 billion. OpenAI itself expects to spend $5 billion this year on computing costs.

This massive investment in custom silicon and infrastructure signals a crucial shift in OpenAI’s strategy. While most AI companies remain dependent on Nvidia’s chips and cloud providers’ data centers, OpenAI is making an ambitious play for technological independence. It’s a risky bet that could either drain the company’s resources or give it an insurmountable advantage in the AI arms race.

By controlling its own chip destiny, OpenAI could potentially cut its computing costs in half by 2026. More importantly, custom chips optimized specifically for GPT models could enable capabilities that aren’t possible with general-purpose AI processors. This vertical integration — from chips to models to consumer products— mirrors the playbook that helped Apple dominate smartphones.

The new search features will appear on ChatGPT’s website and mobile apps. Enterprise customers and educational users will get access in the next few weeks, followed by a gradual rollout to OpenAI’s millions of free users.

For now, Google remains the dominant force in search. But as AI technology improves and more users grow comfortable with conversational interfaces, the competition for how we find information online appears poised for its biggest shake-up in decades.

Autonomous vehicle technology startup Aurora Innovation is targeting April 2025 for commercial deployment of its autonomous trucks, pushing its timeline back by about a quarter. The company had originally planned to launch by the end of 2024. The company said it delayed the launch so it can continue to validate its self-driving technology. 

“While this is modestly later than we had intended, this timing remains within the margin of error we have anticipated and conveyed throughout 2024,” Aurora CEO and co-founder Chris Urmson wrote in his third-quarter earnings shareholder letter. “With our intention to introduce the Aurora Driver with a crawl, walk, run approach, this shift to our timeline will have a negligible financial impact.”

Aurora will go to market as a carrier, but its end goal is to pursue a driver-as-a-service model, wherein carriers purchase trucks with the Aurora Driver tech on board and then offer their services via those trucks to shippers. 

One of the ways Aurora measures the performance and commercial readiness of its Aurora Driver is its use of on-site support, which the company says will be the most expensive support provided. As of the end of the third quarter, the Aurora Driver was delivering commercial loads without the support of a remote human 80% of the time, which is up from 75% in the second quarter. The goal is to reach 90% by commercial launch in the spring. 

The startup intends to deploy up to 10 driverless trucks during commercial launch, with the goal of increasing to tens of trucks by the end of 2025.

Aurora has been testing commercial loads with pilot customers including FedEx, Werner, Schneider, Hirschbach, Uber Freight, and others. The company schedules roughly 160 commercial loads per week, which Aurora says is more than double the volume from last year. As of October 27, 2024, Aurora’s trucks have autonomously delivered more than 8,200 loads and driven over 2.2 million commercial miles — but all with a human behind the wheel.

Aurora, a pre-revenue company building pioneer tech, recorded an operating expense of $196 million in the third quarter, including stock-based compensation of $35 million. That’s less than the $212 million it spent in the same period last year, which Aurora says demonstrates its commitment to being frugal on its path to commercialization. 

The startup ended the quarter with $1.4 billion in cash and investments after raising almost half a billion dollars in August, which should give Aurora runway into 2026 and fund its initial stage of scaling and getting to a place of sustainability.

Correction: This story has been updated to reflect that Aurora Driver was delivering commercial loads without a human driver 75% of the time in the second quarter.

“We will not achieve our target of $500 million in revenue for Gaudi in 2024,” CEO Pat Gelsinger just said on the company’s Q3 2024 earnings call today.

Though Intel just launched its recent Gaudi 3 accelerator this past quarter, said Gelsinger, “the overall uptake of Gaudi has been slower than we anticipated as adoption rates were impacted by the product transition from Gaudi 2 to Gaudi 3 and software ease of use.”

Despite the missed goal, Gelsinger says “we remain encouraged by the market available to us. There is clear need for solutions with superior [total cost of ownership] based on open standards, and we are continuing to enhance the Gaudi value proposition.”

Later on the call, Gelsinger seemingly had some sour grapes to share, pointing out how so far, the industry’s huge spend on AI chips has been focused on training AI models in the cloud. “Training is creating the weather model, not using it,” he says, suggesting once again that putting AI into all the chips, not just ones in the cloud, might be more important in the long run.

Intel reported $13.3 billion in revenue in quarterly earnings today, down 6 percent year over year but up compared to last quarter — and losses of a whopping $16.6 billion. But those losses were based on $18.5 billion of impairments and restructuring charges, the cost of Intel’s decision to rework itself for more profitability in the future.

Last quarter it announced a $10 billion cost reduction plan and over 15,000 layoffs, and it’s now detailing some of the structural shifts inside the company too — including moving its edge computing business into the Client Computing Group that generally handles its desktop and laptop chips, and integrating its software teams into the company’s core business units.

Gelsinger says Intel will “focus on fewer projects, with the top priority to be to maximize the value of our x86 franchise across the client, edge, and data center markets.”

Good morning! Let’s play Connections, the NYT’s clever word game that challenges you to group answers in various categories. It can be tough, so read on if you need clues.

What should you do once you’ve finished? Why, play some more word games of course. I’ve also got daily Wordle hints and answers, Strands hints and answers and Quordle hints and answers articles if you need help for those too.

The New York Times has introduced the next title coming to its Games catalog following Wordle’s continued success — and it’s all about math. Digits has players adding, subtracting, multiplying, and dividing numbers. You can play its beta for free online right now. 
In Digits, players are presented with a target number that they need to match. Players are given six numbers and have the ability to add, subtract, multiply, or divide them to get as close to the target as they can. Not every number needs to be used, though, so this game should put your math skills to the test as you combine numbers and try to make the right equations to get as close to the target number as possible.

Players will get a five-star rating if they match the target number exactly, a three-star rating if they get within 10 of the target, and a one-star rating if they can get within 25 of the target number. Currently, players are also able to access five different puzzles with increasingly larger numbers as well.  I solved today’s puzzle and found it to be an enjoyable number-based game that should appeal to inquisitive minds that like puzzle games such as Threes or other The New York Times titles like Wordle and Spelling Bee.
In an article unveiling Digits and detailing The New York Time Games team’s process to game development, The Times says the team will use this free beta to fix bugs and assess if it’s worth moving into a more active development phase “where the game is coded and the designs are finalized.” So play Digits while you can, as The New York Times may move on from the project if it doesn’t get the response it is hoping for. 
Digits’ beta is available to play for free now on The New York Times Games’ website

The latest Qualcomm Snapdragon 8 Elite has integrated UWB hardware. However, not all phones with the flagship chipset will support the feature.

The Qualcomm Snapdragon 8 Elite has built-in support for UWB

Apple allows precise location tracking. In fact, this is what makes Apple’s AirTags really good at locating misplaced belongings. Although it is getting better, Google’s Find My Device network isn’t as good, and the reason is that it lacks hardware support for UWB.

Ultra-wideband, or UWB, is a short-range wireless communication protocol. It enables super-precise location tracking. It is an additional radio frequency apart from Bluetooth and Wi-Fi.

The Snapdragon 8 Elite is one of the first chipsets to have integrated hardware support for UWB. This is because the SoC packs the FastConnect 7900 connectivity platform, which supports Bluetooth, Wi-Fi, and UWB.

Previous iterations of the FastConnect platform lacked UWB. This meant OEMs needed to add a separate UWB module. Needless to say, several device manufacturers chose not to because of costs and additional space requirements.

Starting with the Snapdragon 8 Elite, OEMs do not need to add any extra UWB hardware, confirmed Qualcomm. FastConnect 7900 is a single-chip 6nm solution wherein, “all features and capabilities are delivered as a single chip solution,” boasted the company.

Why won’t some smartphones not support UWB?

Despite the Qualcomm Snapdragon 8 Elite having integrated UWB support, some smartphones with this chipset won’t support the feature. According to Android Authority, the OnePlus 13, the Realme GT7 Pro, and the Xiaomi 15 are powered by Qualcomm’s flagship chipset, but they do not support UWB. Moreover, some of these devices do not even declare the “android.hardware.uwb” flag, which means the installed Android OS believes there’s no UWB hardware onboard.

A Qualcomm representative has stated, “Snapdragon 8 Elite devices can support UWB thanks to FastConnect 7900, but it’s up to the OEM to determine if they want to utilize the feature.” In other words, device manufacturers will ultimately decide if they want to activate UWB or keep it dormant.

One of the most plausible reasons not to activate UWB would be the varying and complex regulations, and the need to secure prior approval. UWB involves sending out radio waves, and device makers would need to go through the calibration, testing, and certification processes for every country. Since UWB isn’t highly popular with Android device users, manufacturers might be trying to minimize regulatory hassles and costs.