0:08
Assassin's Creed Black Flag Resynced Sales Reach 2 Million on Day 1
Ubisoft just announced that Assassin’s Creed Black Flag Resynced moved over two million copies in its first day, a milestone that puts it among the franchise’s top performers. The launch on July 9 saw the game hit the top spot on Twitch and pull nearly a hundred thousand concurrent players on Steam, the highest the series has ever reached on that platform.
The remake has clearly struck a chord with longtime fans, many of whom have been waiting for a fresh take on Edward Kenway’s pirate adventure. The strong sales figure, which Ubisoft actually shared rather than just player counts, signals confidence that the first real remake of the series is resonating well. Critics have been equally enthusiastic, with IGN rating the game nine out of ten and praising how it updates the classic experience without losing its core appeal.
Even with the impressive numbers, the launch hasn’t been entirely smooth sailing. Some players have voiced concerns over microtransactions, and a few reviews on Steam point out that aspect as a disappointment. Still, the overall reception remains positive, and the community’s gratitude is evident in the flood of thank‑you messages from the studio.
In short, Black Flag Resynced is off to a strong start, combining nostalgic freedom with modern polish, and it’s already drawing a massive crowd of both returning sailors and newcomers eager to set sail.
1:29
Quantum error correction can constantly recalibrate a processor
There are some obvious big picture issues that stand between us and useful quantum computing. Issues like whether we can make enough high-quality hardware qubits to connect into the error-corrected logical qubits we need, and how we generate the states needed to perform universal computation on those logical qubits. But there are also many less prominent challenges that will need to be solved before we can perform calculations.
One of those challenges, which only affects some types of hardware, is calibration. For devices we manufacture, like superconducting qubits, there are always subtle variations among individual qubits. (This is not true when we use something like an atom to hold the qubit, but the lasers that control them can drift.) As a result, this hardware is put through a process called calibration, where we test different frequencies and amplitudes of the microwave pulses that control them to find the combination that produces the lowest error rates, and then save those settings for use in calculations.
However, you can't perform the typical calibration process while e you're doing calculations, which means drift becomes an issue for long and complicated algorithms. Google, though, has figured out that it's possible to do calibration using the same data that's used for error correction.
2:47
Quantum error correction can constantly recalibrate a processor
Google found a way to let the error‑correction routine double as a live calibration loop. Instead of pausing the processor to sweep frequencies, the system watches the syndromes that error‑correction already collects and extracts tiny drifts in the control pulses. Those drift signals feed back instantly, nudging the microwave amplitudes and phases back to their sweet spot while the computation keeps running. The trick works because the same error‑detecting measurements contain enough information about the underlying hardware noise, turning a normally idle diagnostic step into a continuous self‑tuning process. This keeps superconducting qubits aligned without ever stopping the algorithm.
3:30
Quantum error correction can constantly recalibrate a processor
Google’s team showed, in a single experimental study, that the data collected for quantum error‑correction can double‑purpose as a continuous calibration signal. By feeding the same syndrome measurements back into the control software, the processor automatically nudges its microwave pulse settings toward the lowest‑error configuration while a computation runs.
The method sidesteps the usual “stop‑and‑tune” routine that forces a quantum chip offline for recalibration. Because the correction cycles already happen thousands of times per second, the extra bookkeeping adds only a modest overhead, and the drift in qubit frequencies is kept in check throughout long algorithms.
In practice this means a superconducting processor can stay near its optimal operating point without interrupting the workload, a small but concrete step toward more reliable, scalable quantum computing.
4:24
Hugging Face’s CEO on why companies are done renting their AI
Clem Delangue, Hugging Face's CEO, says companies are moving away from renting AI models. They're instead opting for open-source solutions, where they can build and customize their own models. This shift is similar to what happened with software development, where companies used to rent software but now prefer to own and control their own tools.
Delangue notes that companies are starting to realize the limitations of renting AI models. They're finding that these models often don't integrate well with their existing systems, and they're also concerned about data security and ownership. By moving to open-source models, companies can have greater control over their AI infrastructure and avoid vendor lock-in.
This trend is reflected in Hugging Face's own growth. The company has become a hub for open-source AI models and datasets, with around half of the Fortune 500 companies using its platform. Delangue sees this as a sign that companies are looking for more flexibility and autonomy in their AI development.
The shift to open-source AI is also driven by the increasing complexity of AI models. As models become more sophisticated, companies need to be able to customize and adapt them to their specific needs. Open-source models allow them to do this, whereas rented models often come with limitations and restrictions.
Delangue believes that this trend will continue, with more companies moving away from renting AI models and towards open-source solutions. This will require a shift in how companies approach AI development, with a greater emphasis on building and customizing their own models.
5:58
Hugging Face’s CEO on why companies are done renting their AI
I just read this thing about Hugging Face's CEO, Clem Delangue, and it's really interesting. Apparently, he's saying that companies are done renting their AI, and it's all because of this open source movement that's taking off. It's like a GitHub for AI, where developers can share and download these open models and datasets, and now it's being used by half the Fortune 500. Delangue's seen this happen over and over again - companies start using these open models, and then they realize they don't need to rent them from someone else anymore. It's like a shift in how they think about their tech infrastructure.
6:37
Trump says he won’t sign ROAD Act, which could automatically become law tonight
The 21st Century ROAD to Housing Act, after months of deliberation and delays, has encountered yet another curveball.
On Friday morning, President Donald Trump confirmed in a Truth Social post that he won’t sign the legislation, although he didn’t say that he would veto it. Assuming that Trump doesn’t veto or sign the bill today, it is set to automatically go into law at midnight Eastern time, a result that many housing stakeholders expect.
If Trump does issue a veto, Congress could have the votes to override it, but it would cause further delays.
“I will not sign the Housing Bill, which has been fully approved by Congress and sent to the White House, in PROTEST over the fact that the United States Senate is not capable of passing THE SAVE AMERICA ACT…”, Trump said in the post.
The news comes after Trump delayed a signing ceremony for the bill on June 24, instead insisting that Congress first pass the SAVE America Act, a bill aimed at strengthening voter identification and registration requirements. In the days after the cancellation, Trump downplayed the significance of the bill, calling it a “big yawn” in comparison to the SAVE America Act.
Mike Johnson (R-La.), the speaker of the House of Representatives, formally sent the bipartisan housing package — which passed the House on June 23 by a margin of 358-32 — back to the White House on June 29. That triggered a 10-day deadline for Trump to sign the bill, veto it or allow it to become law without his signature. With Sundays excluded, the countdown will end at midnight Saturday after Friday’s deadline passes.
Housing industry stakeholders, who have almost unanimously backed the legislation, are eagerly awaiting its passage.
“The bipartisan 21st Century ROAD to Housing Act is a landmark step toward protecting the American Dream of homeownership. By expanding our housing supply and removing barriers to ownership, this legislation will help more Americans achieve their dream, strengthen communities and build generational wealth,” said Colin Allen, executive director of the American Property Owners Alliance.
The housing package effectively bans institutional investors that already own 350 or more single-family homes from purchasing additional single-family properties. But the final version removed a pair of controversial provisions that largely froze new investments into build-to-rent (BTR) projects.
One of the excluded provisions, which was included in a previous Senate version of the bill, would have included an institutional investor ban without carve-outs for BTR communities. Another would have mandated that new BTR communities be sold to individual homeowners within seven years of completion.
Both proposals, which were ultimately excluded from the final bill, would have significantly undermined the ability of BTR developers to generate returns on their investments.
The final text aligns with Trump’s executive order from January aimed at limiting institutional homebuying, while eliminating the broader restrictions on build-to-rent that raised significant concerns and opposition among housing industry stakeholders.
The 21st Century ROAD to Housing Act aims to streamline the development of housing, primarily by cutting red tape.
For example, the bill would exempt new categories of relatively small-scale development under the HOME program from review under the National Environmental Policy Act of 1969 (NEPA). The legislation would also limit duplicative environmental reviews in the HOME program and make other adjustments to reduce red tape associated with NEPA reviews.
Another provision aimed at removing the permanent chassis requirement from manufactured homes has generated a lot of buzz in the industry. Manufactured housing is an attainable source of housing for millions of Americans, but it is primarily located in rural areas far from city centers. Removing the chassis rule could lower costs while expanding design flexibility, unlocking new opportunities for manufactured housing in higher-cost, infill and urban markets.
Shawn King, executive vice president of national sales and co-founder of Arrive Home, called the bill “the most consequential piece of manufactured housing policy in decades.”
“In the past, federal rules have forced builders to permanently attach a steel chassis to every manufactured home, even though fewer than 7% of these homes are ever moved after they’re installed,” King said. “That requirement alone has been adding $5,000 to $10,000 to the cost of every single home for no real benefit to the homeowner.
“Eliminating it doesn’t just lower the price tag; it opens the door to basements, multi-story designs and layouts that let manufactured homes fit naturally into neighborhoods instead of standing apart from them,” King added.
The legislation also creates grant programs to help state, local and tribal governments update regulatory processes and improve permitting capacity.
11:11
Hugging Face’s CEO on why companies are done renting their AI
Open source AI is booming, according to Hugging Face CEO Clem Delangue. The company has grown into something like a GitHub for AI in recent years, where AI builders can share and download open models and datasets, now used by roughly half the Fortune 500. Delangue has seen the same story play out again and again: companies start […]
11:36
Trump Says He Won’t Sign Bipartisan Housing Bill
President Donald Trump freaked out and said he will not sign a bipartisan housing bill “in PROTEST” against Republicans’ failure to pass his legislation overhauling elections.
Speaker Mike Johnson sent the bill, passed with bipartisan support, to the president’s desk despite Trump blindsiding GOP leaders on the day of the signing ceremony by canceling the event at the last minute.
The move to send the housing bill to the president’s desk set the clock ticking for the legislation to become law with or without his signature, but Trump said he won’t sign it in a lengthy post on Friday.
“I will not sign the Housing Bill, which has been fully approved by Congress and sent to the White House, in PROTEST over the fact that the United States Senate is not capable of passing THE SAVE AMERICA ACT, which is polling at 97% with the Republican Party, and very high with the non-politician Dumocrats,” Trump wrote in a Truth Social post at 9:16 a.m. ET.
If the president does not sign the bill, it becomes law anyway ten days after it is sent to his desk. The president did not say whether he would veto it, but the legislation, which passed with overwhelming support, has a veto-proof majority.
This story is developing and will be updated...
12:49
Trump says he won’t sign housing bill in protest of SAVE America Act inaction
President Trump said on Friday that he will not sign the 21st Century ROAD to Housing Act, which is set to pass into law tonight, in protest over the Senate not passing the SAVE America Act. “THE SAVE AMERICA ACT’S non-passage is CRAZY, and a serious threat to any politician who votes against it!” Trump…
