BANNED USER: temperamental dysregulation
DaseindustriesLtd
late version of a small language model
Tell me about it.
User ID: 745
Banned by: @Amadan
DaseindustriesLtd
late version of a small language model
24d ago
Sol 6 is not going to be launching cyberattacks
A model not much stronger on cyber than Sol 6 Daybreak was responsible for the Huggingface incident.
6.1 Sol is stronger, let alone Astra or 'internal'. If the US moves to strike, then it will be with the very best they can produce. Still good odds the US doesn't strike or makes some insane blunder tbh.
Good odds that the "strike" will simply fail. Also that it's ruled out for non-insane reasons like obvious retaliation.
In WW2, did the Japanese win with their clever infantry tactics and ferociously tough, courageous warriors? Did Germany's excellent staff officers and superior manoeuvre warfare
Risible poetry, drop it and reduce your stim dose. American culture was so superior that in 1949 their managerial technology seemed like alien science to Europeans. Russians also turned out to not be subhumans, to a certain German surprise.
Huawei production is too little, too weak and too late. Nvidia is making about 25x more compute, which clever networking cannot make up for.
Nvidia isn't going to be making 25x more compute, and Americans are not capable of deploying 25x more compute, Nvidia or not. On gigawatts, Semianalysis estimates they have 24GW now and 30GW in the pipeline. On FLOPS it might be more like 10:1 in the near future. Mind you, much of the grandiose American buildout depends on Chinese approval. Americans won't even complete datacenter buildings for 2028 without imports.
Huawei is the largest Chinese vendor, not the only one.
When Deepseek is at the top of the leaderboards, then I'm clearly provably wrong.
That also makes no sense. The volume of inference, as per your logic, is decisive at least as much as peak capability.
What is the failure standard for your prediction/expectation, if Deepseek stays out of the frontier for a certain period of time?
Prediction about what? My prediction is that I am pretty uncertain but I suspect that in he limit of agentic AI, cyber hands the advantage to defenders, so it doesn't matter if DeepSeek or any other Chinese group is exactly at the frontier. It's not clear to me how much of a gap to the actual frontier becomes fatal; and how this translates to compute. My uncertainty interval goes from 1:3 to 1:30 in total compute. I'd say 1:15 is a toss-up.
Now, if you don't mind, drop this condescending poetic act and explain to me: where did you get this part:
Deepseek is stuck in a deadly spiral because they lack compute. Because they lack compute, they struggle to generate much revenue and therefore can't raise enough money to get more data or even acquire more compute.
? Did you write this knowing they have raised $7.4B in June and ≈$15B recently? That Liang's opinion as of June was that they are bottlenecked by the amount of compute accessible for purchase, not by capital?
Per artificial analysis, the new flash is still below Grok 4.7 albeit only slightly
Yes, and it's interesting to look at the actual composition of that score. A year ago, you said:
But Grok 4 just crushes with sheer size I think. It has this 'in this essay I will' style that lmarena certainly isn't going to like, or any normal person really. But it has that heft, it was made for ferociously unsexy mathematics, physics, engineering, research tasks rather than creative writing or coding.
There is one AA benchmark for this, namely CritPt. Look at how Grok 4.7 does there, vs V4.1F.
I'm not making a normative claim here, only a factual/forecasting claim.
I really don't care about your preferences.
DaseindustriesLtd
late version of a small language model
24d ago
Better to be right about something simple than sophisticated and wrong (or be sophisticated and correct about secondary factors).
The question is which factors are primary and which secondary. I posit that one simple factor primary factor is just that DeepSeek's guys have higher IQ and better company culture, and operate in a more materially productive society, which has effects until AI is truly superhuman.
Deepseek is stuck in a deadly spiral because they lack compute. Because they lack compute, they struggle to generate much revenue and therefore can't raise enough money to get more data or even acquire more compute.
Right now DeepSeek can generate that $1B ARR off approximately 3000 Blackwells, and they've locked most of Huawei's production for 2027. They'll be fine on revenue. They have raised about $20B in total so far, and will struggle to spend it on compute anyway.
Regarding the "hack chyna" stratagem, might be of interest, Grok with your "bitter lesson" makes a cameo too. Note that this is Daybreak, so it's not about refusals:
DaseindustriesLtd
late version of a small language model
25d ago
The absolute value is what matters most, there is no prize for being much cheaper.
There is: deployment volume. Though China is disadvantaged on the compute production anyway.
generally, I feel that you're kind of dumb if you feel proud of this "more compute better" discovery. Do you want a cookie? It's trivial. If anything, I'm saying that it's not crushing hard enough. At this rate, your hack-all-China idea likely won't work.
If compute isn't the key factor, then what's holding Deepseek back?
lots of things are holding them back, such as the volume of acquired data and RL environments. Anthropic has been spending on this for longer, and much more. But yes, compute is the most relevant form of capital.
DaseindustriesLtd
late version of a small language model
26d ago
Facebook and Google are back in the game and well ahead of any Chinese competitor.
It's hilarious that we're treating Google DeepMind as some inept nouveau riche that can only "crush" with compute. They invented move than half of this whole field.
Regardless, Grok 4.7 (and you were very confident about Grok) is clearly utterly non-competitive with the latest DeepSeek, and I'm not sure if Muse Wonder and Argon fare much better.
It's a bit unexpected even for me.
My emotions are not so much Deepseek partisanship as cold contempt for you personally, your ritalin-addled writing style and attitude, you irritate me about as much as self_made_human, so I'm using grok to repeatedly mock your year-old prediction driven by faith in Musk successfully brute forcing problems. You take enormous pride in this confident "brute force wins!!" thesis. (Wins what? Well, everything!). A few years ago you were similarly bullish on "China wins by sheer mass" or whatever. I think you just fetishize this idea, megalomania is almost sexual to you. Big datacenters, big models, big rockets, big … but I'll stop myself here. Your arguments for outcomes of either asymmetry have been and are handwaving about qualitative gaps, or sloppy analogies.
I didn't "return" to it, I maintain it. World's cyberdefenses run on indians who ship d = 4 in certificates, hopes and prayers. We didn't have a cyberdefense worth speaking of until this year.
Well, the way things are going, we'll see soon, I guess? By the end of 2027, really mid-2027, AstraFable ought to be more than capable of finding weaknesses in pretty much anything written by humans, and from that point it's a mere question of cost of the attempt, a decision that can be made at will. Though you may be disappointed. Optimistically, it'll audit that stack and find all vulns before AstraFable, which is a bit easier when you have the source code.
Realistically, Americans will pussy out of trying due to the normal fear of retaliation.
More options
Context Copy link