There's a new AI player in town,????? ?????? ?????? and you might want to pay attention to this one.
On Monday, Chinese artificial intelligence company DeepSeek launched a new, open-source large language model called DeepSeek R1.
According to DeepSeek, R1 wins over other popular LLMs (large language models) such as OpenAI in several important benchmarks, and it's especially good with mathematical, coding, and reasoning tasks.
This Tweet is currently unavailable. It might be loading or has been removed.
DeepSeek R1 is actually a refinement of DeepSeek R1 Zero, which is an LLM that was trained without a conventionally used method called supervised fine-tuning. This made it very capable in certain tasks, but as DeepSeek itself puts it, Zero had "poor readability and language mixing." Enter R1, which fixes these issues by incorporating "multi-stage training and cold-start data" before it was trained with reinforcement learning.
Arcane technical language aside (the details are online if you're interested), there are several key things you should know about DeepSeek R1. First, it's open source, meaning it's up for scrutiny from experts, which should alleviate concerns about privacy and security. Second, it's free to use as a web app, while API access is very cheap ($0.14 for one million input tokens, compared to OpenAI's $7.5 for its most powerful reasoning model, o1).
SEE ALSO: OpenAI could release agentic AI tool Operator soonMost importantly, this thing is very, very capable. To test it out, I immediately threw it into deep waters, asking it to code a fairly complex web app which needed to parse publicly available data, and create a dynamic website with travel and weather information for tourists. Amazingly, DeepSeek produced completely acceptable HTML code right away, and was able to further refine the site based on my input while improving and optimizing the code on its own along the way.
I'll do all of that...tomorrow. Credit: Stan Schroeder / Mashable / DeepSeek I also asked it to improve my chess skills in five minutes, to which it replied with a number of neatly organized and very useful tips (my chess skills did not improve, but only because I was too lazy to actually go through with DeepSeek's suggestions).
I then asked DeepSeek to prove how smart it is in exactly three sentences. Bad move by me, as I, the human, am not nearly smart enough to verify or even fully understand any of the three sentences. Notice, in the screenshot below, that you can see DeepSeek's "thought process" as it figures out the answer, which is perhaps even more fascinating than the answer itself.
We get it, you're smart. Credit: Stan Schroeder / Mashable / DeepSeek It's impressive to use. But as ZDnet noted, in the background of all this are training costs which are orders of magnitude lower than for some competing models, as well as chips which aren't as powerful as the chips that are on disposal for U.S. AI companies. DeepSeek thus shows that extremely clever AI with reasoning ability doesn't have to be extremely expensive to train — or to use.
Topics Artificial Intelligence DeepSeek
New Cast, New Energy
Trump seemingly tweets conspiracy theory about bombs sent to his critics
15 dog toys that your pooch (probably) can't destroy
Trump seemingly tweets conspiracy theory about bombs sent to his critics
On the Road with Hello Kitty
Twitter shames Trump for doing the absolute least in the wake of explosive devices
This very accurate costume of The Rock's turtleneck outfit adds one perfect detail
Twitter shames Trump for doing the absolute least in the wake of explosive devices
‘Mind Game’ at the Egyptian
Heidi Klum just outdid herself (again) with her Shrek
Celebrate the Year of the Dog with JACA and AKIHO
Let's start calling the Russian 'troll' attack what it really is
The anatomy of the niche teen Instagram meme
Congressman roasted for his boring choice of Halloween treats
English Rakugo Performance in Torrance
This stray cat who crashed a fashion show is my favorite supermodel
9 of the best celebrity trolls
This election night Photoshop battle imagines wild political agendas
‘Free! Take Your Marks’ Screening Nationwide for One Night Only
The Instagram artist who turns pop divas into 'Pokémon' leaders warms my gaymer heart
Trump adds 'Apprentice' villain Omarosa to White House staff5 Days of Awesome Wallpapers: Cars and Sports WallpapersUber gives cities a peek under the hood, sharing trip data with new toolGuys show us their best poses for Tinder profile photosRuth Bader Ginsberg pens inspiring handwritten note to 8Channing Tatum gets social with Jenna Dewan's NSFW nap timeYou can take the subway, but you can't hide from Donald Trump's faceLonely Planet gives quokka hot spot the recognition it deservesSeveral Indian colleges ban sexist singers from performing on campusMark Hamill reads Trump tweet in his creepiest Joker voice and nails it Joe Biden gives surprisingly candid response to protester at Clinton rally Potter fans battle it out over J.K. Rowling Albus Potter tweet Refusal to remove cowboy hat is totally worth the suspension Parent gives brutally honest excuse for tardy kids Hangry little hedgehog will only cheer up if you give him treats Uber riders can now use pre Mom and toddler's 'potty time' song is way too catchy Can you find the hot dogs among the Instagrams of people's legs? I kicked the crap out of Dyson’s new vacuums and they wouldn’t fall over Social senior dog walks 4 miles every day to catch up with all his friends
0.1336s , 9927.359375 kb
Copyright © 2025 Powered by 【????? ?????? ??????】China's DeepSeek AI might be smarter than OpenAI's smartest AI,Feature Flash
