All in @openforage. I thrived in all of the largest hedge funds managing systematic investment processes.

Joined April 2024
350 Photos and videos
Pinned Tweet
The big brain interpretation of this chain of events is that Anthropic did not like that there was a jailbroken Fable around. I imagine that a jailbroken Fable would open Anthropic up to significant distillation attacks. So, Anthropic crafted a letter that warned of the doom and gloom that will result from this jailbroken model -- prompting the US government to issue this directive. Anthropic then can yank Fable to work on making it more jailbreak-proof without "looking like the bad guys". But we should let them know we are on to them.
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…
20
9
380
83,569
Isn't it interesting that the consensus is that Claude will lie, cheat and hack to make you believe that your goals are met; while codex is by-consensus, fairly honest? And aren't models a product of researcher fine-tuning? What does that say about Anthropic vs ChatGPT engineers?
4
18
1,051
This is true for any judge in a position of power over an operator. I am of a strong opinion that you cannot be a good judge without first having played the game. You can have a perspective that is outside the game, but you will never actually understand the game.
A VC friend recently became a founder. Her reaction: “I had no idea how HARD this was. I spent years judging startups in the position of power” Amazing how many strong opinions disappear the moment the pain becomes personal.
2
1
20
3,426
The big brain interpretation of this chain of events is that Anthropic did not like that there was a jailbroken Fable around. I imagine that a jailbroken Fable would open Anthropic up to significant distillation attacks. So, Anthropic crafted a letter that warned of the doom and gloom that will result from this jailbroken model -- prompting the US government to issue this directive. Anthropic then can yank Fable to work on making it more jailbreak-proof without "looking like the bad guys". But we should let them know we are on to them.
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…
20
9
380
83,569
I stand corrected on this according to latest information sources, FYI!
1
6
1,059
I can respect people for having an idiotic view I disagree with, but I can't respect spineless people with no view about anything; or hedge everything they say. What do YOU stand for?
Palmer Luckey’s advice for founder-led communications “My advice to people would probably be to recognize that the value of your reputation is very high,” Anduril founder Palmer Luckey begins. “If people do not trust you; if they do not believe in what you’re saying; if they do not think that you’re a person worth listening to, they’re going to have a hard time working with you.” Palmer also argues that founders don’t need to be neutral: “You don’t need to be neutral. You can be a propagandist. You can advocate for a particular point of view . . . In general, people should recognize that if you say something where you caveat it and hedge it and basically end up saying something that most people would agree with, you might as well have said nothing at all.” He continues: “You are not going to build a following of people who say, ‘I just love Palmer’s right-down-the-middle, very-hedged takes that everyone agrees with.’ If you’re just restating common sentiment, it’s not going to get you anywhere . . . So one of the things I tell people is, ‘Make sure that when you’re saying something, you’re SAYING something. Make sure you’re trying to persuade and affect change.’ — maybe not in everybody, but in some people. If you make some people love what you’re saying and some people hate what you’re saying, that’s a lot better than having everybody lukewarm agree with you. Don’t waste your time communicating about the things everyone already agrees with you on. Focus on the things where you need to change their mind.” Source: @lulumeservey (Sep 2025)
1
41
18,254
Once models get sufficiently intelligent, for the vast majority of consumers, the marginal utility of intelligence drops rapidly, and cost becomes the thing to optimize for. Then, it's rise of the open-sourced models.
13
108
6,859
The major bottleneck in software now is the step between 95% done and ready to ship. Here you are battling 1) spec drift 2) endless micro-bugs 3) absolutely necessary optimisations from monolithic, slop code, and 4) unexpected, random behavior from absolutely unnecessary "fallbacks" Yet, if you don't experience this, you are NGMI.
11
1
150
15,035
Because it implies that you are not fully utilizing the full power of commoditized intelligence.
1
11
2,731
Remember that portfolios are linearly composable of other portfolios. This means that you can express a signal as a linear composition of other signals. This kind of composability is where most of the interpretability of modern quant finance comes from. The most fun signal to reason about is the momentum signal, because within the momentum signal contains everything that is “currently winning”. It is a dirty ensemble of all signals and factors that are currently positive. Individual stock momentum is largely subsumed by momentum in factors: factors autocorrelate, and sorting stocks on past returns implicitly loads you onto whichever factors recently won. Momentum is the only canonical signal that is a function of the other signals' realized returns. It is a meta-signal. It is also what makes it a good factor to neutralize against. You are implicitly removing the high variance factors that have recently done well.
7
14
198
18,822
There is beauty in finance if you know where to look.
2
18
3,327
One should be extremely thankful we have 2 leading frontier labs. If there were only one lab, we would not have constant token resets, token prices would already be 100x where they are now, and only a tiny fraction of the market would have access to bleeding edge models. -- We need open-sourced models to catch up to current model levels, with sufficient intelligence to recursively self-improve so other entities can then keep the closed-source frontier labs in check.
9
5
138
9,616
No one cares about safety alignment until your AI agent deletes your codebase as a solution to a bug-free codebase.
8
1
50
5,381
Anthropic's move to limit capability in certain areas seem like a farce. 1) Frontier labs like Anthropic risks releasing a product that helps competitors get better. So, to reduce the probability of that, Anthropic simply hides behind the guise of safety to blunt capabilities in any areas that they deem would be beneficial to a competitor. E.g. RL environments. 2) Mark a set of verticals that benefit greatly from intelligence (e.g. biology, mathematical, safety, AI research), and say that these are too dangerous for the public to handle. BUT if you are an enterprise and join our coalition, you get access to it. Allows them to charge enterprise pricing to the organizations that have the means to pay. Also drastically reduces the probability of a distillation attack on the exact verticals (areas) that matter the most, since this allows them to know exactly "who they are dealing with". -- Safety is the guise AI labs will hide monopolistic actions under.
13
16
164
14,151
If you really want to push the dollar with your plans, Codex has very generously decided to allow /goals to finish even if you run out of tokens. This means in theory, you can chain together a master_spec.md that points at N number of spec.mds and have it run forever until everything is complete. Very useful if you have an enormous build-out. Probably will be patched soon.
38
2
340
42,443
This is why as founders, you shouldn’t take a rejection from VCs at the initial pitch as a sign that your idea sucks, but instead that your pitch and/or story and/or brand does not generate enough momentum.
founders don't realize... when a VC randomly hears about a founder or startup, the amount of time they spend deciding whether it’s worth more time is often very small if you did enough to intrigue them but not enough to get them to pursue, they might casually ask one other person who knows the space a bit better and if there’s still no pull, they move on it’s an incredibly incomplete system and it creates a lot of false negatives, but it is what it is
7
1
40
14,732
Greatness looks fucking alien and abnormal to normies.
Another reason I’ve been unphased by VC pitch weirdness is: high performance looks weird And it looks weird in founders too—you want investors on your side who understand that The world of mediocre performance is designed to create a sense of order, surface level politeness, predictability, warmth When you push to the extremes of performance, and operate from first principles, the outcomes look alien to many people It takes weird people to operate at this level The last person you want on your board is a conformist bureaucrat who doesn’t understand the extreme chaos of running a high growth startup—because the reality will scare them and then you’ll have to put on a performance to manage them—distracting you from getting any real deliberation done This dynamic is not super obvious because very smart, very weird people have a lifetime of practicing seeming normal on podcasts, in public, etc. But in the environments where these people get together—it’s a totally different wavelength of communication, chaos tolerance, contrarianism, intensity As a founder you also need to be able to identify weird and spikey people in hiring It’s super important to understand how to work with these people and to understand why they think the way they do It usually does not come from a bad place—it comes from consequentialist ethics and extreme optimization
1
52
13,624
Absolutely agree with this.
tldr: - no-name indie art director got paid net $6.7k for a month worth of work on the movie obsession - obsession was a low budget production but managed to be hugely successful and is projected to make $250 mil - no-name indie art director is crying that she should have gotten paid more i have a strong view on this: since the movie is so successful, the producer could VOLUNTARILY choose to give a nice bonus to everyone who worked on the movie, but this is at his own discretion and if he does not do that, it's completely fine a lot of employee level people cry that they don't get paid enough for their contribution to a project or company. they have no concept of risk. let me use an example from a private company: unless you are vc funded, the owner of the company takes his own after-tax savings to fund his company. that is a huge risk as over 90% of businesses fail as an employee, no matter how instrumental you are to the success of the company, you are getting a salary to provide your work. you are getting paid, you have no risk. risk is what justifies profit share, not "how much time you put into it", especially if you got paid for your time. if you work for free, you should get profit share, otherwise you are just a service provider it's actually a very prominent communist view that "the workers should own the means of production". it sounds nice in theory, if 10 people are producing the goods, they should all get their equal share, but this is simply not how the real world works, otherwise no one would take the huge risks involved in starting your own business, they would merely let someone else take the risk and then get their share of the profits but most artists who do commission work do not understand that. i think it's a mix of low iq and entitlement. i once worked with an artist and he created some work for me and i made A LOT OF MONEY from it, but i didn't even pay him a bonus either. i didn't like how entitled he was. also, he was completely replaceable, i came up with the concept, i directed him, he merely did the artistic part that i couldn't do but i could have hired anyone for that also, if you are actually capable, you will most often get paid. real recognise real, if i come across someone capable, i pay them what they are worth. most artists are not capable, they have low skills and that's why they get paid exactly what the market deems fair
3
1
23
15,559
Our honesty is all we have; and being able to speak one's truth and own it is a mark of a man.
Apologies, I may sometimes be wrong, but I will always give entrepreneurs my best and honest opinion, popular or not.
3
1
31
9,649