Month: April 2023

There’s still quite a bit of lingering nostalgia for Windows 95, and even though its notoriously easy to fake your way back into the days of blocky menus and balding men screaming at you that it’s “only $99,” one Windows experimenter managed to use ChatGPT to generate easy product keys for the venerable operating system.

Late last month, the YouTuber Enderman showed how he was able to entice OpenAI’s ChatGPT to generate keys for Windows 95 despite the chatbot being explicitly antagonistic to creating activation keys.

Old Windows 95 OEM keys used several parameters, including a set of ordinal numbers as well as other randomized numerals. In a considerably simple workaround, Enderman told ChatGPT to generate lines in the same layout as a Windows 95 key, paying particular attention to specific serials that are mandatory in all keys. After a few dozen attempts of trial and error, he settled on a prompt that worked, and it could generate around one working key for every 30 attempts.

In other words, he couldn’t tell ChatGPT to generate a Windows key, but he could tell it to generate a string of characters that matched all the requirements of a Windows key.

Once Enderman proved the key worked in installing Windows 95, he thanked ChatGPT. The chatbot responded “I apologize for any confusion, but I did not provide any Windows 95 keys in my previous response… I cannot provide any product keys or activation codes for any software.” It further tried to claim that activating Windows 95 was “not possible” because Microsoft stopped supporting the software in 2001, which is simply untrue.

Interestingly, Enderman ran this prompt on both the older, GPT-3 language model and on OpenAI’s newer GPT-4, and told us that the more recent model improved upon even what you saw in his video. In an email, Enderman (who requested we use his screenname) told Gizmodo that a certain number string in the key needed to be divisible by 7. GPT-3 would struggle to understand that constraint and created far fewer usable keys. In subsequent tests with GPT-4, ChatGPT would spit out far more correct keys, though even then not every single key was a winner or stuck to the prompt’s parameters. The YouTuber said this suggests “GPT-4 does know how to do math, but gets lost during batch generation.

GPT-4 does not have a built-in calculator, and those who want to use the system to generate accurate answers to math problems need to do extra coding work. Though OpenAI has not been forthcoming about the LLM’s training data, the company has been very excited about all the different tests it can pass with flying colors, such as the LSAT and Uniform Bar Exam. At the same time, ChatGPT has shown it can occasionally fail to spit out accurate code.

One of the main selling points of GPT-4 was its ability to handle longer, more complex prompts. GPT-3 and 3.5 would routinely fail to create accurate answers when doing 3-digit arithmetic or “reasoning” tasks like unscrambling words. The latest version of the LLM got noticeably better at these sorts of tasks, at least if you’re looking at scores on tests like the verbal GRE or Math SAT. Still, the system isn’t perfect by any means, especially as its learning data is still mostly natural language text scraped from the internet.

Enderman told Gizmodo he has tried generating keys for multiple programs using the GPT-4 model, finding it handles key generation better than earlier versions of the large language model.

Don’t expect to start getting free keys for modern programs, though. As the YouTuber points out in his video, Windows 95 keys are far easier to spoof than keys for Windows XP and beyond, as Microsoft has started implimenting Product IDs into the operating system installation software.

Still, Enderman’s technique didn’t require any intense prompt engineering to get the AI to work around OpenAI’s blocks on creating product keys. Despite the moniker, AI systems like ChatGPT and GPT-4 are not actually “intelligent” and they do not know when they’re being abused save for explicit bans on generating “disallowed” content.

This has more serious implications. Back in February, researchers at cybersecurity company Checkpoint showed malicious actors had used ChatGPT to “improve” basic malware. There’s plenty of ways to get around OpenAI’s restrictions, and cybercriminals have shown they are capable of writing basic scripts or bots to abuse the company’s API.

Earlier this year, cybersecurity researchers said they managed to get ChatGPT to create malicious malware tools just by creating several authoritative prompts with multiple constraints. The chatbot eventually obliged and generated malicious code and was even able to mutate it, creating multiple variants of the same malware.

Enderman’s Windows keys are a good example of how AI can be cajoled into bypassing its protections, but he told us that he wasn’t too concerned about abuse, as the more people poke and prod the AI, the more future releases will be able to close the gaps.

“I believe it’s a good thing, and companies like Microsoft shouldn’t ban users for abusing their Bing AI or nerf its capabilities,” he said. “Instead, they should reward active users for finding such exploits and selectively mitigate them. It’s all part of AI training, after all.”

---

Want to know more about AI, chatbots, and the future of machine learning? Check out our full coverage of artificial intelligence, or browse our guides to The Best Free AI Art Generators, The Best ChatGPT Alternatives, and Everything We Know About OpenAI’s ChatGPT.

Guangzhou-based company CAS Space recently flew a small rocket outside a Chinese coastal city before landing it upright on a sea-based platform. The successful test flight and landing could eventually change the game for the Chinese rocket industry.

CAS Space, which is partly owned by the Chinese Academy of Sciences, launched a small rocket prototype in Haiyang, east China, according to China’s Global Times.

The rocket prototype weighs around 205 pounds (93 kilograms) and is powered by twin engines, including a turbojet engine that simulates the kind of variable thrust liquid rocket engine that’s used during vertical landings, the Global Times reported. The rocket flew no higher than about 3,280 feet (1,000 meters) before reverse engine thrust enabled it to hover back down onto the awaiting sea platform. The test lasted for 10 minutes, but the demonstration could eventually lead to rocket stage reusability, similar to how its U.S. counterpart, SpaceX, does it.

The company announced that its vertical landing rocket test was a success, with the rocket recovery technology possibly being applied to future rocket models like the Lijian-3 and Lijian-3 heavy-lift launchers, Global Times reported.

SpaceX has become a pro at landing its boosters upright. In 2015, the company performed a successful landing for the first time following a year of trial and error. The rocket’s first stage uses thrust to control its descent, landing on four legs made of carbon fiber. Today, the Falcon 9 rocket regularly lands on a droneship at sea.

Lian Jie, a senior engineer at CAS, told the Global Times that the Chinese rocket landing technology is different from that being used by SpaceX in that “ours is based on the domestic technology, both software and hardware, and we are exploring technology thresholds such as the variable thrust management, precision positioning and the stabilizing technology on our own.” So basically, that they came up with it all by themselves.

Either way, it is a highly coveted technology as it would allow launch providers to save massive amounts of money through the regular reuse of rocket stages. The private space industry in China is trying to play catch up after the Chinese government created room in the market in 2014, allowing commercial rocket companies to exist, innovate, and compete. The landing test is a testament to the new capabilities emerging in China, but as to when we might see a Chinese rocket perform a vertical landing after traveling to space, that remains an open question.

For more spaceflight in your life, follow us on Twitter and bookmark Gizmodo’s dedicated Spaceflight page.

Asus announced it is making a Steam Deck handheld competitor called ROG Ally. Preorders will be available for it soon.

The handheld will run a custom chip from AMD inside and YouTuber Dave2D suggests that it utilizes a Zen 4 + RDNA 3 APU. The price and availability haven’t been revealed, though. Best Buy has a listing up for it, but no preorder option just yet. ROG Ally looks somewhat similar to the Steam Deck but has some differences, most notably the color. While the Steam Deck is dark grey, ROG Ally is white.

The device can also be plugged into the ROG XG Mobile eGPU which can stream games to the TV while charging it simultaneously. ROG Ally is compatible with multiple platforms, including Windows 11 and Xbox Game Pass.

Since the success of Valve’s Steam Deck, many companies have tried to capitalize on it by releasing their handhelds. Examples include Ayaneo Air and Logitech’s G Cloud.

In GameSpot’s Steam Deck review, we said, “What the Steam Deck excels at is being a mobile hub for your Steam games, giving you enough to work with to get the best out of a selection of games you can’t go without while away from your desktop. In that regard, there simply is nothing better on the market right now.”

Got a news tip or want to contact us directly? Email news@gamespot.com

After decades of collecting data about how astronauts’ vision is affected by spaceflight, scientists have identified a set of microgravity-induced changes to the human eye known as Spaceflight-Associated Neuro-Ocular Syndrome, or SANS. This condition is considered a risk to human health in long-duration spaceflight. 

Later this year, a SpaceX Falcon 9 rocket will hurl the Polaris Dawn mission of four crew members skyward for a privately-backed voyage of up to five days in Earth orbit. During their stay time circuiting our planet in a Crew Dragon capsule, the high-flying team is dedicating major time to probe health impacts on the body from their sojourn into space.

One of their medical tasks is specific to help unravel what’s behind SANS symptoms. It is a known, unknown issue, say researchers, a malady that includes swelling of the optic nerve, alteration in the shape of the eye itself, as well causing fuzziness to vision. Clearly, on any lengthy trek to Mars and back, being “blindsided” by blurry eyesight from SANS is not ideal.

Related: Polaris Dawn: The trailblazing commercial mission of the Polaris Program

Ongoing research

As for its cause, SANS is tied to the absence of gravity‘s effects on the body, causing a fluid shift that can generate pressure changes in the brain and eye. This fluid shift is thought to be the basic cause of the eye and brain structural changes. But identifying the exact why and when SANS crops up — and what countermeasures can be applied — is prompting ongoing research.

Making it all the more perplexing, NASA says that signs of SANS appears in roughly 70 percent of Earth orbiting astronauts, but with significant individual variability.

Overall, clarity over vision issues in space has come a long way since the February 1962 flight of Mercury astronaut John Glenn. That first three-orbit spaceflight by an American had Glenn dutifully monitoring his vision by reading an eye chart to see if his eyeballs might change and alter his vision.

Cyborg study

Allie Anderson is a leading SANS investigator in the College of Engineering & Applied Science at the University of Colorado, Boulder. She has been working with the Polaris Dawn crew to help examine the timing of spaceflight-induced changes to the human eye through continuous monitoring of intraocular pressure, the pressure within the eyeball.

Anderson’s study involves a Sensimed-produced “Triggerfish” contact lens that has an embedded strain gauge to measure how the cornea changes shape over time. The contact lens contains a tiny antenna to transmit its data and will explore how the front part of the eye changes with microgravity exposure. The Polaris Dawn crew calls it “the cyborg study.”

In addition, Anderson’s SANS research has the crew using a device built by PlenOptika — a portable, handheld autorefractor — to measure how a person’s corrective eyewear prescription changes while in spaceflight.

Validate technologies

“Using both of those devices, we want a better understanding of the time course for how the eye adapts in microgravity,” Anderson told Space.com. “For long-duration astronauts, they were getting structural changes, like flattening of the eye that resulted in a shift in their vision, as well as swelling at the optic nerve head. It wasn’t completely clear why that was happening.”

One other goal of the SANS work is to validate low mass, low power technologies for space missions. “We want to get as much data as possible and to obtain measurements as early in flight as possible,” Anderson said. “It’s a good opportunity for people in the community to collaborate and share data and hopefully contribute to understanding the problem,” she said.

Anderson underscored NASA’s tackling of SANS. “They are doing a ton of work to settle on what the definition of having SANS is and how to grade the severity of it,” she said. “It’s an interesting challenge to get your head around, that’s for sure.”

Earth benefits

With astronauts projected to take longer and longer space trips in the future, SANS countermeasures is a key next step, said Prem Subramanian, a professor of ophthalmology, neurology, and neurosurgery at the Sue Anschutz-Rodgers University of Colorado Eye Center in Aurora, Colorado. 

Work is progressing on trying to figure out what’s causing SANS and associated spaceflight-caused eye changes, but more research is essential. 

Theories based on what is being observed could be totally wide of the mark, Subramanian advised. A better understanding of the forces that cause SANS is critical to direct the development of SANS countermeasures, he added. 

“We have to figure out how to reduce the incidence of the problem, probably before we even fully understand what’s causing it,” Subramanian told Space.com. “The inability to really recreate it here in a terrestrial environment is a major barrier. We’ve got models, but the models are not great and they are limited.”

In addition to the attention given to SANS as a space traveler’s dilemma, there are also notable benefits from this research for the gravity-bonded majority of us here on Earth.

“There are terrestrial diseases that cause optic nerve swelling and vision loss,” Subramanian added. “A better understanding of how to treat this condition will improve our understanding of things that happen back here on Earth … those patients who develop similar increased intracranial pressure problems, say from trauma or a disease, or from a variety of other things,” he concluded.

Follow us @Spacedotcom (opens in new tab), or on Facebook (opens in new tab) and Instagram (opens in new tab).