DeepSeek’s $6M AI Model Outperforms OpenAI’s Costlier Offerings

by

2023 saw the emergence of DeepSeek, an innovative Chinese AI company established by Liang Wengfen, capturing attention in tech circles. Their latest creation, the R1 AI model, runs on V3 technology, and according to tests, outperforms OpenAI’s o1 reasoning model in areas like math, science, and coding. What sets it apart even more is that the model was developed for a reportedly small investment of $6 million compared to the costs typically associated with proprietary models.

In response to the growing excitement surrounding DeepSeek, influential figures within the industry have been expressing their opinions on this groundbreaking technology across various media outlets and social platforms. Among them is Sam Altman, CEO of OpenAI, who has also voiced his thoughts on the matter. According to Altman:

The DeepSeek R1 model stands out, offering remarkable performance given its cost. Rest assured, we’ll develop even more advanced models. It’s exciting to see a fresh competitor enter the scene! Let’s plan for some upcoming product launches.

DeepSeek’s R1 model is indeed remarkable, especially considering its performance relative to the cost. Rest assured, our offerings will surpass their quality significantly. It’s truly exhilarating to see a new competitor emerge! Let’s prepare for some exciting announcements coming up.

(January 28, 2025)

Read moreApple Maps now works on your Windows PC... sort of

In the year 2023, I witnessed Arnaud Bertrand contrast DeepSeek’s accolades with his own focus on OpenAI’s executive remarks following a ‘Conversations’ presentation to Indian venture capitalists. During this event, Sam Altman appeared to downplay competition from budding AI startups, particularly those grappling with financial constraints of approximately $10 million, as reported by Tom’s Hardware.

According to Sam Altman:

As an analyst, let me clarify the situation: It seems that we have a significant advantage in terms of our foundation model training. I strongly advise against attempting to compete with us on this front, as I genuinely believe it’s a challenging endeavor. However, I also acknowledge that your role requires you to strive for success, even in difficult circumstances. In essence, I am expressing my conviction that the task might be quite daunting.

It’s undeniable that the progress in generative AI comes with a substantial price tag. Last year, Sam Altman shared an ambitious vision of AI development that would require approximately $7 trillion and several years to construct 36 semiconductor facilities and additional data centers. This bold proposal reportedly earned him the nickname “podcasting bro” from TSMC executives.

Read moreRemnant 2 announces third and final DLC 'The Dark Horizon' on its birthday, and its release date isn't far off

It’s no secret that advancements in generative AI are expensive. Last year, Sam Altman proposed an ambitious plan to spend $7 trillion and construct 36 semiconductor plants and data centers over several years. This idea led some TSMC executives to call him a “podcasting bro.

Earlier this week, OpenAI and SoftBank officially announced a massive $500 billion project called Stargate, aimed at building data centers to support cutting-edge artificial intelligence developments. Just last year, there were whispers that suggested OpenAI might be on the brink of bankruptcy, predicting losses up to $5 billion within the next 12 months.

As a result, the creators of ChatGPT have rolled out a new $200 monthly Pro Plan. This plan, as stated by Altman, is designed to tackle complex problems more effectively. It appears that OpenAI is introducing ChatGPT Pro in an effort to expand its income sources during challenging financial times.

Read More

2025-01-29 23:41