@digitalages

Complex Litigation @DocketRunner

 Florida, USA
Joined April 2010
  • Tweets463
  • Following1.4K
  • Followers110
20Media
Polymarket
@Polymarket8h
JUST IN: Anthropic recorded a staggering $80,000,000,000.00 net loss in the first half of 2026.
21
JRF retweeted
JRF
@digitalagesOct 19
Replying to @Sony91121271
Law firms give up their unique expertise to Anthropic The next model becomes an average of all the expertise Anthropic sells the average output back to the law firms
1
2
85
I wonder if we can detect this with something simple like asking the person to summarize what they just said
Elon Musk
@elonmuskOct 19
So many people are going to get one-shotted by super persuasive super intelligences and I might be one of them
38
In my experience working with LLMs, this approach tends to bias the output generating less impressive and creative outputs LLMs are more likely to generate a genuinely clever strategy when they’re free to be wrong, just like us
Zhiyuan Ren
@ZhiyuanRen8Oct 19
Replying to @digitalages @LegalDataHunter
True, but 'it depends' is literally dependent on the specific facts/constraints—which is context. Experience isn’t magic intuition; it’s knowing which contextual signals actually matter and how to weigh them.
40
Isn’t it true that the labs are incentivized to make their harnesses require more human input because our prompts are their data? We may have competing interests wrt automation.
30
JRF retweeted
JRF
@digitalagesOct 16
Replying to @Sony91121271
We should talk I’m a practicing litigator building tools inside my own firm My company @DocketRunner plugs directly into Claude/ChatGPT to make law firms more AI native across their tools with minimal behavioral change for legal assistants, paras and attorneys
1
2
279
I was ready to write Dot off but I guess it’s back in the game now
1
36
I love that lawyers are embracing AI but branding as 'agentic law firm' today is about as stupid as "email law firm" in 1995...this just the new normal and barely any kind of edge. Maybe gives a slight marketing boost for some subset of prospective clients for the next 6 months but equally some prospective clients will be skeptical you're shipping slop, so it's a wash.
26
4
89
4,476
Who are the ai-pilled complex litigators around here?
1
38
Hello world! Interactive, independent benchmarking for Legal-AI Try it out ⬇️📊
4
1
2
34
5,831
Machines won’t take over until a printer works when you need it
djcows
@djcowsOct 15
AI will never end humanity because the agents will get stuck in a loop building tests before they can even launch the first nuke or bioweapon
40
Claude desktop has become beyond unusable inside of Windows
49
If your first internet access came through a CD sent in the mail…
Sush
@2sushSep 25
If your first email ended in @ yahoo, @hotmail, or @msn... you have seniority on the internet.
227
my mother-in-law is now on instinct i agreed to join her trusted agent network we’re still so early
84
hey @noahrshinn my wife wants to be part of instinct’s GTM for moms who should she speak to?
39
update: i invited my wife to instinct
JRF
@digitalagesSep 3
Anecdote for anyone building consumer AI: my wife became obsessed with using AI in the last three months. She uses Gemini inside Google search. She has still never downloaded the Gemini app or ChatGPT.
101
the main reason this is more exciting than Claude for Legal is bc Codex is a much better harness also Anthropic models have slumped recently maybe the new integrated claude/cowork app can bring it back above water
OpenAI
@OpenAISep 17
Astra for Law: Frontier intelligence built for your practice. A new offering powered by GPT-6 Astra with tools, settings, and context to support the expertise and judgment of lawyers and legal technology firms.
2
144
i think the new meta is realizing that using frontier intelligence for classification and retrieval may not be worth the cost
Sherwood
@shcallawaySep 17
I guess the new meta is figuring out how to reframe your entire product, technology, business, and industry as a collection of classification problems
2
124
i’m starting to think @typesafeai isn’t real and the waitlist is a psyop
2
3
128
interesting
Diogo Almeida
@CompleteSkepticSep 15
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x cheaper (w/ output tokens free) • Frontier composable intelligence optimized for decisions AFAICT the shortest path to AI-based economic revolution
92
Load more