Saturday , June 21 2025
$30

US scientists claim to replicate DeepSeek for $30 dubbed “TinyZero,”

Researchers at the University of California, Berkeley, claims they’ve managed to reproduce the core technology behind DeepSeek’s at a total cost of roughly $30.

The news raises questions about whether developing advanced AI requires huge budgets or if cheaper alternatives have been ignored by major tech companies. DeepSeek recently launched R1, an AI model that aims to replicate ChatGPT and similar systems at a significantly lower training cost.

Russia detects first SuperCard malware attacks via NFC

Russian cybersecurity experts discovered the first local data theft attacks using a modified version of legitimate near field communication (NFC)...
Read More
Russia detects first SuperCard malware attacks via NFC

Income Property Investments exposes 170,000+ Individuals record

Cybersecurity researcher Jeremiah Fowler discovered an unsecured database with 170,360 records belonging to a real estate company. It contained personal...
Read More
Income Property Investments exposes 170,000+ Individuals record

ALERT (CVE: 2023-28771)
Zyxel Firewalls Under Attack via CVE-2023-28771 by 244 IPs

GreyNoise found attempts to exploit CVE-2023-28771, a vulnerability in Zyxel's IKE affecting UDP port 500. The attack centers around CVE-2023-28771,...
Read More
ALERT (CVE: 2023-28771)  Zyxel Firewalls Under Attack via CVE-2023-28771 by 244 IPs

CISA Flags Active Exploits in Apple iOS and TP-Link Routers

The U.S. Cybersecurity and Infrastructure Security Agency (CISA) has recently included two high-risk vulnerabilities in its Known Exploited Vulnerabilities (KEV)...
Read More
CISA Flags Active Exploits in Apple iOS and TP-Link Routers

10K Records Allegedly from Mac Cloud Provider’s Customers Leaked Online

SafetyDetectives’ Cybersecurity Team discovered a public post on a clear web forum in which a threat actor claimed to have...
Read More
10K Records Allegedly from Mac Cloud Provider’s Customers Leaked Online

Canada 2nd largest airlines “WestJet” investigates cyberattack disrupting internal systems

WestJet, Canada's second-largest airline, is looking into a cyberattack that has affected some internal systems during its response to the...
Read More
Canada 2nd largest airlines “WestJet” investigates cyberattack disrupting internal systems

Paraguay 7.4 Million Citizen Records Leaked on Dark Web

Resecurity found 7.4 million records of Paraguayan citizens' personal information leaked on the dark web today. Last week, cybercriminals attempted...
Read More
Paraguay 7.4 Million Citizen Records Leaked on Dark Web

High-Severity Flaw in HashiCorp Nomad Allows Privilege Escalation

HashiCorp has revealed a critical vulnerability in its Nomad tool that may let attackers gain higher privileges by misusing the...
Read More
High-Severity Flaw in HashiCorp Nomad Allows Privilege Escalation

SoftBank: Over 137,000 personal info leaked

SoftBank has disclosed that personal information of more than 137,000 mobile subscribers—covering names, addresses, and phone numbers—might have been leaked...
Read More
SoftBank: Over 137,000 personal info leaked

Alert
Trend Micro Apex One Flaw Allow Attackers to Inject Malicious Code

Serious security vulnerabilities in Trend Micro Apex One could allow attackers to inject malicious code and elevate their privileges within...
Read More
Alert  Trend Micro Apex One Flaw Allow Attackers to Inject Malicious Code

The Berkeley team’s response? To do it even more cheaply. Led by PhD candidate Jiayi Pan, the researchers created a smaller-scale version, dubbed “TinyZero,” and released it on GitHub for public experimentation. Though it lacks the massive 671-billion-parameter heft of DeepSeek’s main offering, Pan says TinyZero captures the core behaviors seen in DeepSeek’s so-called “R1-Zero” model.

Pan’s approach uses reinforcement learning, where AI begins with random guesses and improves its answers by exploring different solutions. He mentioned the Countdown game, a British TV puzzle that challenges players to combine numbers to reach a target. Pan noted that while the AI initially produced inaccurate results, it eventually learned to correct its mistakes.

The bigger revelation:

The notion that just a few days of work and a small amount of money can achieve a key AI capability surprises many in the industry. This challenges the common belief that significant AI advancements need large data centers, expensive GPUs, and massive funding.

DeepSeek made waves in the tech industry by claiming that training its main model costs only a few million dollars, much lower than many U.S. companies spend on AI. Pan and his team stated it can be done for just $30 on a small scale.

Skeptics urge caution, noting that DeepSeek’s reported affordability may not reflect the whole truth, as the company could be using alternative resources or proprietary distillation techniques.

Pan’s “TinyZero” demonstrates that advanced reinforcement learning is possible on a budget, but it may not match the range of tasks that the larger DeepSeek system can perform. TinyZero might be more of a basic proof-of-concept than a serious competitor.

The demonstration suggests a significant change in the AI landscape. If open-source developers can create advanced features with limited resources, it questions why major companies like OpenAI, Google, and Microsoft spend so much on their platforms. While developing scale and advanced capabilities is expensive, it raises concerns about potential cost inflation in the industry. Open-source projects could potentially compete with these tech giants by working with smaller budgets.

“TinyZero” and DeepSeek’s R1 indicate a growing appetite for compact, resource-friendly AI models. Many had assumed that cutting-edge breakthroughs required billions in expenditure. Now, it seems a bright graduate student or a scrappy startup might surprise the world with innovation on the cheap. Whether this ultimately reshapes the future of AI infrastructure or remains an exciting anomaly, the conversation around affordable, powerful AI is only beginning.

Full experiment log: https://wandb.ai/jiayipan/TinyZero
GitHub: https://github.com/Jiayi-Pan/TinyZero

Source: interestingengineering.com

ChatGPT, DeepSeek, Qwen 2.5-VL Vulnerable to AI Jailbreaks

Check Also

FortiGate

Hackers retain access to patched FortiGate VPNs using symlinks

Recent incidents continue to bring this into focus with active exploitations of known vulnerabilities as …

Leave a Reply

Your email address will not be published. Required fields are marked *