alexwwang
alexwwang on Hacker News
Don’t you find the official document website was out of service for a long time since the new model was published soon?
Posted by
alexwwang on Wednesday August 12, 2026 @01:47PM
from the self-post dept.
Its benchmarks are comparable to Fable 5, with each having its strengths and weaknesses, resulting in a near tie.Terminal Bench 2.1: 72.1 → 87.9, surpassing Opus 4.8's 85.0, and only 0.1 away from Fable 5's 88.0.Cybergym: 52.7 → 83.3, surpassing Opus 4.8's 78.3, and slightly higher than Fable 5's 83.1.DeepSWE: 12.8 → 62.7, surpassing Opus 4.8's 58.0.AutomationBench: 12.8 → 31.8, surpassing Opus...
54M tok is really not a big number, tbh. Hurry up dude, consume 54B before the boots touch down.
It's just a hypothesis, not proved.
I might be the similar cat of user with you. But I have to say by now it push me to burn more token with it. No one could predict what will happen tomorrow and how high the price would bump to.
Interesting insights. I will consider these factors. AFAIK, the electricity price is lower in China than in US.
Show me if you could. I am interested in your answer.
I agree you on these issues. When we look back to the first and second industry revolution, how many companies survive to today since then? It takes time and it’s a relay game, but more fast than past.
So I think there’re huge opportunities for everyone involving in.
Talking back to deepseek’s announcement, I think it is smart and funny but we needn’t to be panic on this forecasted price rising at all. Just sit and relax, and wait for big news announced by other model players. At least glm 5.3 is on its way.
I agree that the trend of token price should be going down but it should also consider the effect and gain per token so it varies among different models, mainly according to their capability and availability. There’s a long way to go before it meets the equilibrium state.
So I think we should not limit our field of view on software development but on realms where ai software has not penetrated yet. That’s the blue ocean for ai business. The forth industry revolution hasn’t begun yet. The new steam engine powered textile machine is on the stage of developing and experiment yet.
Posted by
alexwwang on Thursday August 06, 2026 @01:07AM
from the self-post dept.
My take: this notice from DeepSeek might actually be a brilliant marketing move. The logic here is to urge users to ramp up their usage over the next 2 to 3 months. By the time the price hike actually kicks in, newer models will likely have already drawn users away, and whenever DeepSeek drops its next, even stronger model, the new price point won't feel nearly as steep. It’s a clever way to...
The peasants’ war happened over and over again in Chinese history among the gaps of dynasties. That’s a result of autocracy.
Posted by
alexwwang on Sunday July 12, 2026 @08:43PM
from the news.mit.edu dept.
Posted by
alexwwang on Thursday July 09, 2026 @11:05PM
from the blog.chuanxilu.net dept.
This is a significant work. LLM might be not only a model of language but a model of a part of brain.
The situation has been changed. Google is no longer the one it was.
Posted by
alexwwang on Wednesday June 24, 2026 @05:29AM
from the blog.chuanxilu.net dept.
That makes sense. I haven’t found it work in api yet.
Does the api support vision yet?
I usually install Lazarus on my pc though seldom use it now. Still want to pick it up someday to compensate my miss in childhood. Only used turbo C then.
If you know what Chinese are suffering mentally, you may understand why I say so. Criticize a model is not the smart way to against a system.
I think maybe it’s a tool and it’s up to you to make use of tools to try to let more Chinese people know and convince them to believe your idea. Don’t blame a tool but make proper use of it to make a better world.
It has been available now in coding plan.
I wonder if these local model could really solve problems especially for users that aren’t experts on a given coding language. I am not sure that, more than inline auto completion and unit implementation, are these model capable of designing and composing tech specs that really work.
I knew it at once after you said. But politicians don’t fully understand what they face.
Yes. But they are minority.
Yes. It’s really not a good idea to make this ban. When the US is gradually isolated in this way by its gov’s policy, the world becomes more and more dangerous. What worse, the traditional value of open to competition that Americans have hold for centuries seems to be substituted step by step. It’s absolutely a tragedy.
People always exaggerate the thing they don’t understand.
I hope so. But how? Who gonna fund these projects and how to coordinate with every sides. This is complex. I only believe that the open source AI won’t lack users.
This, IMO, is another scenario. MCP is designed and played as a part of the automatic tool chains. These are two different types of needs. But in the case you mentioned, when some parts of the work should be automated, it’s also possible to utilize mcp there.
I agree. Mcp might be useless in a personal scenario but it absolutely plays a role of service infrastructure in organizations. It is another form of api for those abilities that are not wrapped with rest api yet. But when they are wrapped in mcp, it seems not necessary to wrap them into rest api or cli again in near future. So these mcp services survive.
The only thing matters is how to import these mcp services into agent context on demand or say by the gradual disclosure principle.