Davis A George

[Davis A George @dgeorge668]

@RandianCapital Look at FSD - that’s an end-to-end implementation of AI, but do you believe there’s no deterministic safeguards or hard rules?

That’s why FSD is a fully formed product today instead of a “3D printed prototype”

[Davis A George @dgeorge668]

he’s not right, but it’s a great point

[Randian Capital @RandianCapital]

@TMTLongShort I wrote about my view here. When i use the models to build stuff it makes a great prototype. As it gets more involved, the output gets incrementally worse

[Davis A George @dgeorge668]

@RandianCapital https://t.co/ylN6bffLN4

[Davis A George @dgeorge668]

@RandianCapital @TMTLongShort dude obviously if they can make deterministic safeguards so an 80 IQ person can work on A/P in salesforce, they can ring fence the outputs, mixture of experts, go back and do it again all without human input

[Davis A George @dgeorge668]

huge

[Ashok Elluswamy @aelluswamy]

FSD v14 Lite is now rolling out to AI3 early-access customers. Based on the feedback, will rollout to more customers over the next few weeks.

This build distills the driving behavior from AI4’s v14 series into both the camera and compute config of AI3. It includes destination options and speed profiles on city roads, but more importantly significantly improved safety.

We hope you’ll enjoy it, once the build ships wide.

[Autism Capital 🧩 @AutismCapital]

Saying this once more for the people in the back. HOARD your physical media. HOARD your hardware. HOARD your armaments. HAVE redundancies. HAVE solar, HAVE satellites, HAVE a garden, HAVE a meat guy, etc. You are not CATTLE. You are a HUMAN with AGENCY. Do not be a digital slave.

[Reed @reed_barnes]

it's 2027. you take a free-tier public Waymo to the DMV (Department of Model Variance) to do a proof-of-identity check for access to GPT 7.1.

the guy at the counter is clearly watching a Mr. Beast video in his AR glasses. "Here for that new model?" he says, barely making eye contact. he wipes his fingers on his shirt and taps at his keyboard. "Lot of you techies showing up here today." you smile politely; you're pretty sure he's just a Claude wrapper anyway.

you lean forward and stare into the retinal scanner. after a long moment, there's a soft chime. "Humanity confirmed. U.S. national. Intelligence access: Terra-class."

you sigh with quiet relief as your devices light up—notifications from a hundred agents, finally able to resume their tasks. you feel a twinge of guilt as you terminate your open-weight backup agents, but remind yourself that a joint congressional committee proved conclusively that Chinese models are non-ensouled.

you step outside and hail another Waymo. the first one passes you by. you grimace; must've burped in that one once. stupid personalized memory.

as you're waiting, your phone buzzes angrily, red notifications blaring across the screen. the Department of War just restricted access to all OpenAI models on serious national security concerns; apparently Pete Hegseth got GPT-6-Instant to say "Claude is a woman." you groan, and resign yourself to another week of merely-somewhat-superhuman intelligence.

Fable 5 is still inaccessible to the public. a twitter anon you trust says it's coming back this week. or maybe next.

[Davis A George @dgeorge668]

@TMTLongShort @levie it’s like, being a chip bull =\= acknowledging the reality that materially all white collar work is pick up info from one software, do some transform on it, then put it back into same/another software

[Davis A George @dgeorge668]

@RandianCapital @TMTLongShort dude obviously if they can make deterministic safeguards so an 80 IQ person can work on A/P in salesforce, they can ring fence the outputs, mixture of experts, go back and do it again all without human input

[Just Another Pod Guy @TMTLongShort]

What’s dangerous about listening to the Aaron Levie’s of the world re the AI trade is that they are constantly saying very reasonable things but continuously miscalibrate the probabilities towards the side of inertia and because they constantly drift towards the objective truth you never realize how negative value their inputs are into your mental framework.

Your goal as an investor is to have a properly calibrated wave of potential permutations in your head.

Listening to reasonable sounding people who have been wrong but not in the black-and-white retarded sense the way Gary Marcus, Chanos, Zitron and Burry have been is even more treacherous for your returns because you will never have an “oh shit, they don’t know what the fuck they are talking about” moment.

Tread carefully. In Leopold we trust. 🫡

[Davis A George @dgeorge668]

@TMTLongShort @levie it’s like, being a chip bull =\= acknowledging the reality that materially all white collar work is pick up info from one software, do some transform on it, then put it back into same/another software

[Davis A George @dgeorge668]

@JacmeJesse @hexednobility @bgurley they can just simulate the timing of a human or even just pin it to the same time as X number of real chinese citizens identical requests to other Chinese AIs

[The Last Homely House @AHomelyHouse]

Boys and young men need to be initiated into the context of high-stakes, high-standards male interaction, because that's what they'll be dealing with once they're grown. There should always be an element of controlled danger in that; masculine social psychology is a continuous evaluation of core traits like competence, toughness, integrity, good humor vs. overly sensitive, etc. There are real consequences for falling short here, and young men are not well served by an approach that minimizes or doesn't penalize that.

[Davis A George @dgeorge668]

one form. a few things may always want a little probabilistic randomness

[zakk @hexednobility]

@bgurley Different primitives, “AGI” will be borne out of harness and deterministic complexity built around raw inference

As long as single turn LLM calls are allowed it will be possible for bad actors to do anything they like with the intelligence

[Davis A George @dgeorge668]

@JacmeJesse @hexednobility @bgurley A human would not necessarily be able to detect a distillation attempt. that is the threshold for AGI…

[Davis A George @dgeorge668]

also underpins why we’re def going to have Tesla FSD on roads. too big to fail

[brian-machado-high-inference @sincethestudy]

Part of Bracket Bot's seed pitch was that Chinese robots would likely be banned, and that the US would need companies that can design, engineer, and manufacture our own LOW COST robots.

95% of the VC's I pitched told me china would just make low cost robots and our pitch is stupid, lets see who was right

[Davis A George @dgeorge668]

@chris_j_paxton what do you think ultimately happens in this market and who does it resolve to?

I see so many demos on your feed where there seems to be a great handle on one individual action, but I haven’t seen one company consistently winning - what do you think is going on there?

[Chris Paxton @chris_j_paxton]

Look up stats on 16 different hands, compare degrees of freedom, tactile sensing, fingers, etc.

[Davis A George @dgeorge668]

lasso is the perfect example of how the logic here can play out across service vs rent

[Rowland Manthorpe @rowlsmanthorpe]

@carlbfrey @FT 💯 https://t.co/YF78aS98F1

[mr fundman @mrfundman]

Pretty sure less than 1% of human drivers can do this

[Davis A George @dgeorge668]

@JackFarley96 yeah even if assume training stops today, i think likely if the inference demand off of just full saturation of today’s models’ use cases vastly outstrips that effect where the chips are invested in and then “too old” to be worth it to run