Joined August 2009
254 Photos and videos
Derek Colley retweeted
Keep it simple if you want to learn local ai 1) Build the cheapest rig that can house 1 used 3090 2) Who cares about ram, cpu, etc, buy ddr4 32gb, get a decent mid range ryzen/intel 3) download qwen 3.6 27b You’ll be into the rig like $1500 at most You won’t be able to upgrade much (maybe another 3090) but it’s a great starting point You don’t feel locked into learning about big rigs and upgrade paths If you like it then you can still keep and run the 3090 rig (trust me it’s useful) And you can build a new separate rig if you need more vram, on any platform after you have some experience and know what type of hardware you actually need, not what people recommend online!
37
3
192
17,327
Derek Colley retweeted
i like the terminal, but a sidebar is... nice say hi to Apple Pi!🍎small mac app for pi agent full Swift, 3 MB of download, no Electron (🖕) projects left | sessions middle | terminal tabs right oh, it also has a remote SSH mode! open source: github.com/dodo-reach/apple-…
2
4
33
2,152
Derek Colley retweeted
Replying to @DerekColley_
What open source AI?
1
1
1
21
Derek Colley retweeted
Jun 13
If you have the NVMe Go download as many models as you think you might ever want. Now, go on Huggingface. They’re coming for open models next.
123
85
1,287
79,656
So, even with the best-in-class models and all the compute they have, their strategy was still to not make a deal… did they even use AI reasoning to thrash this out?!
A short history of how we got here, because the chronology is the whole story. January: the Pentagon demands unrestricted use of Claude for autonomous weapons and domestic surveillance. Anthropic says no. February: the President orders every federal agency to drop Anthropic. The Defense Secretary bans Pentagon contractors from doing business with them. A rival announces its classified-network deal within hours. March: the Pentagon designates an American company a "supply chain risk" under a statute written for foreign adversaries. A federal judge blocks it. May: the Pentagon signs AI deals with seven companies. Anthropic is not one of them. June 9: Anthropic releases Fable 5. June 12: Commerce issues an export control directive over a jailbreak that, by the government's own account, was demonstrated verbally, came with no written explanation, and involves a capability you can get from other publicly available models today. Two things are true at once. First: Anthropic spent months marketing Mythos as too dangerous to release. Sam Altman said it was "incredible marketing to say we have built a bomb." The Commerce Department has now formally agreed it is a bomb. If you describe your product as a munition in every press release, eventually a government takes you at your word. They wrote the legal predicate themselves and called it a brand. Second: we have run this experiment before. In the 90s the government classified encryption as a munition under ITAR. Activists defeated it by printing PGP's source code as a book, because books are protected speech and floppy disks were arms exports. A t-shirt with three lines of RSA Perl was legally a munition. The controls collapsed because math does not stop at customs. The new wrinkle is the "deemed export" rule: showing controlled technology to a foreign national inside the US counts as exporting it abroad. Which is why Anthropic's own foreign-national employees are now locked out of the model they built. The munition is in the building and the people who made it are not allowed to look at it. The jailbreak is the paperwork. The refusal was in January.
24
Ordered a replacement lid from @_FrankGreen. - arrived without a silicon seal... bloody useless! Anyone know where I can get a replacement seal?
10
It's like a Ford Mustang and Jaguar XF had a baby. I just think it's not sporty or sleek enough... Side view looks like Jaguar X-type. What other cars can you see in the DNA?
Where classic design meets quiet confidence Peugeot 504 Coupe
47
Derek Colley retweeted
M3 would never 🙂‍↔️ As a matter of fact, the weights are now open, too. huggingface.co/MiniMaxAI/Min…
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Claude models is not affected. We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible. Read our full statement: anthropic.com/news/fable-myt…
245
477
6,607
515,958
Derek Colley retweeted
What people are missing about the US government's AI regulation announcement: ID verification will now be forced on all accounts to prove citizenship. Frontier labs will take your data, and your sovereignty is officially dead. A permanent underclass division and a total control society are beginning right now. People ignored me when I started saying this last year, but it is happening right in front of our eyes. Get into Open-Source and Sovereign AI. Advancing together through collective intelligence is the only way to fight back.
As a result of a US government directive, we are suspending access to Claude Fable 5 for all users. You can continue to use all other Claude models. Here’s what this means for you: Across Claude products, new sessions will run on your selected default model or Opus 4.8, and existing Fable 5 sessions will end with an error. On the Claude Platform, requests to Fable 5 will also return an error. Please update your integrations to other Claude models. We know this is a disruption to your workflows; we appreciate your patience and support.
180
351
1,944
302,133
Derek Colley retweeted
Jun 13
Kimi finally released a desktop app, and it has Swarms! Now I'm stuck using K2.6 but I'm sure they're upgrade it. It comes with: - swarm mode - plugins index - chat/build modes - browser extension for browser use! Clean design from Kimi team. Let's see if they catch up
20
14
402
16,881
Derek Colley retweeted
Jun 12
History's first trillionaire is a guy who catches rockets out of the sky with chopsticks and beams internet to every dead zone on the planet. Same guy ships cars that drive themselves, humanoid robots for the factory floor, brain chips that let paralyzed people move a cursor with pure thought, and an AI running on a supercomputer his team stood up in months instead of years. And the people crashing out about his net worth are doing it on the app he owns. The same app governments spent years trying to censor. You cannot legislate a rocket into orbit.
1,859
11,488
70,277
2,300,223
This is why I push opensource AI. Self-sovereign: money, compute and intelligence. In the next months, local AI will be 'good-enough'. Will it be able to catch up? Dunno, not sure if AI is exponential or sigmoid.
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Claude models is not affected. We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible. Read our full statement: anthropic.com/news/fable-myt…
1
1
44
Derek Colley retweeted
I've seen him discuss it in a few interviews youtu.be/ThApNwas6ps Other people have said that it's a solvable problem, possibly within 20 years.
1
1
1
48
If you met @elonmusk, and you could ask one question, what would it be? It's unlikely that I will meet Elon in person, so I'll ask my one question here: Where does personal longevity feature in the list of things that need to be solved? You don't seem to be working on this - at least publicly. Does that mean it's: - not relevant - can be solved later - can't be solved - other? Asking for a friend... 🤷🏼‍♂️
2
1
2
358
whoa, would could actually have our 1st *real* superhero on earth?
21
Harness > Agent Tourbillon Agents just got: - Workspace: read/write files, knowledge, private & shared - Sandbox: execute code, run tests But the real kicker? They're "just" agents, and we need harness Today I'm implementing @Mastra Code as a plugin harness for Tourbillon. code.mastra.ai/reference
20
/Ask I have long conversations with my code... I ask about options, pros/cons, I ask to go deeper on certain issues. Or I pivot and dig into something the AI found via search, code inspection or just recommended. Sometimes 45 mins, until I'm happy with the direction. /Plan Then I ask the AI to make a plan to implement /Build ...
1
18
Tourbillon is a local-only coordinator for agentic services. Prompt: I want my agents to have optional tool to execute code. AI: hold my beer... ⚡️<beer:electric meme?> Tourbillon agents can now execute code in a local sandbox 🚀 Building on @mastra is sweet...!
1
25
Derek Colley retweeted
Jun 12
same model. same prompt. one flag changed. minicpm5 1b running locally on my dgx spark via vllm. left pane has thinking off, answers in 7 seconds, confidently tells me clock hands overlap at 00:00:65. a time that does not exist lol right pane is the exact same checkpoint with enable_thinking=true. it reasons out loud for 20 seconds and nails it. 1:05:27 most people have no idea hybrid reasoning exists in 1b models. this thing fits in 2gb nothing sped up, raw token stream at 30fps @OpenBMB
3
7
25
3,079