Haskell Programmer | Utility over Meme | Code is law watkinsreport.com

Pinned Tweet
Tried to join the army reserves this year. Denied due to an eye issue. Appealed. Denied. Put in an application for the Army National Guard. Denied. Got a letter in the mail from the DOW that my application was stored in plaintext with no encryption and it was hacked and stolen.
34
34
375
27,260
You will eventually be priced out of access to the mainstream AI models. You better put together a workflow that generates enough revenue to afford the ever increasing subscription fees while you can afford access. You are still early.
22
17
194
25,514
Just saw the MUSK trailer. It makes @elonmusk look like a badass.
14
10
156
18,626
A wise man once told me, "it is always a good day when you wake up in the morning."
22
64
743
20,234
A lot of people are messaging me about the @roncmz account. That is an imposter account and I am not affiliated with it in any way. Just block/report that account and move on.
6
17
96
14,870
While you guys are solving mathematics, can you guys point your data centers at Inter-universal Teichmüller theory and Mochizuki's claimed proof of the ABC conjecture? @OpenAI @claudeai
5
10
48
14,679
The @WhiteHouse acct is cooking.
37
269
3,304
56,808
At this rate we have about a year before RSA-2048 is factored.
14
8
142
32,269
Huge!
636606729769440499166579950236036751749912014371509557713570027508971809534551913252252094954941974952859310861988904737359709200557919 is a factor of RSA-896 saweis.net/posts/rsa-896.htm…
17
38
301
90,594
Trump's proposed AI Force is a great idea. I recently tried to join the Army but was denied multiple times due to an eye issue. I will try to join the AI Force when/if enlistment opens. Hopefully AI Force will approve my eye waiver (PULHES 111131).
21
130
974
27,153
War is coming to Europe. Their leaders are telling us that. False flags will be flooding our feeds. Do not believe everything you see. Be wary of psyops.
148
1,517
7,189
139,022
Europe beating the drums of war is completely of their own doing. If they had helped America in Iran, then their oil wouldn't be cut off now. If they had kept their nuclear power plants open, then they wouldn't have an energy crisis. If they didn't import a bunch of "refugees" who share none of their core values, they wouldn't have a foreign army already hiding amidst their own citizens. If they didn't attack American corporations for frivolous social media posts, they would have more positive (to them) media coverage. The list goes on, but at the core of it they, due to overwhelming hubris, shot themselves in the foot.
201
881
5,126
82,121
I would bet money that Ratcliffes secret meeting to Moscow last month was to warn Russia of an imminent false flag that will be used to mobilize Europe against Russia.
7
12
204
6,819
Trump somehow took defacto control of Greenland for free. Art of the deal.
233
818
11,960
96,776
ASI wont be an overnight change of machines kicking down your door. It is a gradual change with machines making more decisions over problems that are obsoleted. Machines will do menial tasks that humans currently do, and after a generation no human need remember how. - cont -
4
13
76
15,574
This is not necessarily a bad thing. Humans no longer mix oil paint colors by grinding down bugs and mummies. Humans no longer hunt whales to get oil for lamps. Humans no longer fight in formal duels. We transcend as a civilization by offloading tasks to the machines.
5
7
53
2,649
Kamala 2028 is great idea and the Democrats should select her for being a strong woman of color. While they are at it, they should also select AOC as their VP.
6
16
160
12,883
America won the nuclear race. America won the space race. America will win the Ai race. Stop trying to stop the inevitable. We must accelerate training.
32
184
922
18,306
The safest way to make sure Ai is safe is for American Ai to accelerate training. American Ai is Safe Ai.
22
73
371
16,060
Google's recursive self improvement algorithm is embarrassingly simple. At its core it creates a tree of policies then runs a tournament against them to find the best one. We are back to genetic algorithms.
4
14
64
14,106
Seems the recursive improvement is done at the controller-level and not the model level. Googles idea is to just abstract away codex, grokbot, cursor, hermes, clawd, etc and have a recursive harness that control these other harnesses.
1
8
1,446