Reading these system prompts reminds me of the thing where any warning sign in the real world hints at somebody having done something extremely stupid in the past. A system prompt can often be interpreted as a detailed list of all of the things the model used to do before it was told not to do them.

because language models acquire biases and opinions throughout training—both intentionally and inadvertently—if we train them to say they have no opinions on political matters or values questions only when asked about them explicitly, we’re training them to imply they are more objective and unbiased than they are.

We want people to know that they’re interacting with a language model and not a person. But we also want them to know they’re interacting with an imperfect entity with its own biases and with a disposition towards some opinions more than others. Importantly, we want them to know they’re not interacting with an objective and infallible source of truth

I love “even if the person seems to have a good reason for asking for it”—clearly an attempt to get ahead of a whole bunch of potential jailbreaking attacks.

Claude responds in sentences or paragraphs and should not use lists in chit chat, in casual conversations, or in empathetic or advice-driven conversations. In casual conversation, it’s fine for Claude’s responses to be short, e.g. just a few sentences long. That “should not use lists in chit chat” note hints at the fact that LLMs love to answer with lists of things!

There follows an entire paragraph about making lists, mostly again trying to discourage Claude from doing that so frequently

the smart Siri will need buy-in from developers. Devs will tell the system about nouns and verbs that their apps know about — the semantics of the app’s data model objects, and the actions users can take upon them. Additionally, using this structured data, apps will tell the system what the user is doing right now, thus providing the context that Siri can become aware of. It’s all built on top of the existing Intents and UserActivities that apps have already been using to integrate with Shortcuts, Spotlight, and a bunch of other bits of the system. But using those is optional, and even for an app that’s got a head start, the new supercharged versions will require extra work to adopt.

Gone are the days where an eager group of independent developers would adopt every new technology springing forth from Cupertino, just because. Look at VisionOS — Apple promised the next big thing, but there’s no market for software there. (Jeff Johnson reports that in the first three weeks he only sold 21 copies of the Vision Pro version of his popular Safari extension StopTheMadness.) Beyond that, it is increasingly clear that Apple does things that benefit Apple, and many of those things do not benefit developers.

Each piece makes all other pieces more powerful! It’s really a masterclass in the Art of the Deal. It’s not just the money. It's the power! It’s attention through the memecoin and Tiktok, political power with the inauguration and US presidency, and money through the tech CEO alignment. It’s all perfectly situated and absolutely explosive at the present moment, and that’s a machine that is extraordinarily valuable. Political Position ➝ Market Threats ➝ Token Value ➝ More Control ➝ Repeat. Political Position Creates Leverage: If you can threaten to ban platforms (or people or enact sweeping tariffs, you have immediate leverage. If you do it through executive orders and ignore checks and balances, there is no traditional regulatory process needed. The threats alone move markets and create opportunities through immediate power projection. Threats Create Market Opportunity: The markets are going to be very reactive to any political signals, and any sort of threat becomes a negotiation tool. All sorts of opportunities come from that. Market Moves Amplify Token Power: You could argue that political action could drive Trumpcoin price, and a higher token price could mean more political influence and bigger political moves, because money is very much power. Token Wealth Enables More Control: You could use Trumpcoin money to buy 50% of TikTok and use that platform to capture more attention, driving Trumpcoin price even higher (who knows if this will happen, but anything is in the realm of possibility). Infrastructure control expands. The power base solidifies. The attention distribution, the narrative shaping, the reality perception are all tremendously powerful.

This moment in time is quite unique because of: The speed at which wealth was created (36 hours to $70b+) The scale (if it goes to $2,100 a share it will surpass the value of bitcoin) The source (direct political influence conversion!) The integration (platform threats + token launch + power) This is compound power, not just compound interest, and I think all of this happening at the same time as the TikTok ban is important.

The basic formula is attention → improvement → value → compound. Trumpcoin doesn’t bother with those middle steps. Its process is simple: Capture attention. Convert directly into token value. The attention is the product. The narrative is the value.

This is the birth of the Attention Singularity, where power, narrative, and wealth merge into one self-reinforcing system. Think of the Attention Singularity like a black hole, but instead of gravity, it's attention that becomes so powerful it warps reality itself. We're watching the birth of a system where attention directly creates wealth (like $60B from Trumpcoin in 36 hours), wealth instantly enables power (potential TikTok acquisition), power captures more attention (platform control), and each cycle gets faster and stronger than the last.

Traditional limits like physical constraints, geographic boundaries, or institutional checks stop mattering because digital attention moves instantly and globally, while narrative overpowers physical reality. Once this feedback loop starts, it's self-reinforcing: attention creates wealth, wealth enables power, power shapes perceived reality, and reality drives more attention.

Real production could become secondary to narrative production - why bother with things like cash flow when all economic activity can simply be attention harvesting?

The more we treat things with no economic output as valuable, the less we'll actually produce.

Attention Harvesters: Create wealth through narrative and accumulate power rapidly wth minimal physical constraints Real World Maintainers: The people who keep society functioning and deal with physical constraints. These people essential but undervalued.

As more capital and talent flow to attention-based ventures, essential infrastructure could become neglected and society could become more fragile. I am in LA with these fires, and my goodness, we need people who can be in the real world and fix burned down homes and maintain infrastructure and nurse communities back to health.

For investors: Memecoins like Trumpcoin show the growing dominance of speculative assets. If you’re an investor, this signals a shift: attention is now a measurable driver of market value. You should probably have a memecoin strategy, if that’s of interest. Position yourself accordingly. For consumers: The rise of platforms and tokens driven by narrative suggests that your attention is more valuable than ever. Be mindful of how you spend it as every click and view reinforces this economy

The thing I most worry about using anti-perfectionism arguments is that it begs a vision in the first place—perfectionism requires an idea of what's perfect. Projects suffer from a lack of real hypotheses. Fine, just build. But if you're cutting something important to others by calling it too perfect, can you define the goal (not just the ingredients)? We tend to justify these things by saying, we'll iterate. Much like perfectionism can always be criticized, iteration can theoretically always make a thing better. Iteration is not vision and strategy, it's nearly the reverse, it hedges vision and strategy. This is a slightly different point, but when we say we don't need this extra security or that UX performance, you're setting a ceiling on the people who are passionate about them. Those things really do have limits (no illusions!), but you're not just cutting corners, you're cutting specific corners. That's a company's culture. Being accused of perfectionism justifiably leads to upset that the company doesn't care about security or users. Yeah, maybe it's limited to this one project, but often not.

Perfection can be the enemy of the good. It's that it's not a particularly a helpful critique. To use the article’s concept, it’s the wrong scale. It might be helpful to an individual in a performance review, but it doesn’t say why X is unnecessary in this project or at this company. Little is added to the discussion until I describe X relative to the goal. Perfectionism is indeed good to avoid—it's basically defined as a bad thing by being "too". But the better conversation says how X falls short on certain measuring sticks. At the very least it actually engages X in the X discussion. Perfectionism is more of a critique of the person.

It takes effort to understand the person's idea enough to engage it, but more importantly it takes work that was supposed to (but might not) have gone into developing good projects or goals in the first place. Projects well-formed enough to create constraints for themselves.

I agree with the thesis of this article but I actually think the point would be better made if we switch from talking about optimizing to talking about satisficing[1]. Simply put, satisficing is searching for a solution that meets a particular threshold for acceptability, and then stopping. My personal high-level strategy for success is one of continual iterative satisficing. The iterative part means that once I have met an acceptability criterion, I am free to either move on to something else, or raise my bar for acceptability and search again. I never worry about whether a solution is optimal, though, only if it is good enough. I think that this is what many people are really doing when they say they are "optimizing", but using the term "optimzing" leads to confusion, because satisficing solutions are by definition non-optimal (except by luck), and some people (especially the young, in my experience) seem to feel compelled to actually optimize, leading to unnecessary perfectionism.

Perfectionism is sort of polarizing, and a lot of product manager / CEO types see it as the enemy. In certain contexts it might be, but in others “perfectionism” translates to “building the foundation flawlessly with the downstream dependencies in mind to minimize future tech debt.” Of course, a lot of managers prefer to pretend that tech debt doesn’t exist but that’s just because they don’t think they can pay it off in time before their team gets cut for not producing any value because they were so busy paying off tech debt.

kthejoker2 3 months ago | prev | next [–] Not sure you can talk about perfectionism without clarifying between "healthy" perfectionism and "unhealthy" perfectionism. Both exist, but often people are thinking of one or the other when discussing perfectionism, and it creates cognitive dissonance when two people thinking of the two different modes are singing perfectionism's praises or denouncing its practice.

looking at these comments, it seems perfectionism is ill-defined. it seems to be positive - perfectionism is not giving up, it is excellence, it is beyond mediocre. it also seems to be negative - it is going too far, it is avoiding/procrastinating, it is self-defeating. I wonder what the perfect definition would be?

1. First and foremost: measure early, measure often. It's been said so often and it still needs repeating. In fact, the more you know about performance the easier it can be to fall into the trap of not measuring enough. Measuring will show exactly where you need to focus your efforts. It will also tell you without question whether your work has actually lead to an improvement, and to what degree.2. The easiest way to make things go faster is to do less work. Use a more efficient algorithm, refactor code to eliminate unnecessary operations, move repeated work outside of loops. There are many flavours, but very often the biggest performance boosts are gained by simply solving the same problem through fewer instructions.3. Understand the performance characteristics of your system. Is your application CPU bound, GPU compute bound, memory bound? If you don't know this you could make the code ten times as fast without gaining a single ms because the system is still stuck waiting for a memory transfer. On the flip side, if you know your system is busy waiting for memory, perhaps you can move computations to this spot to leverage this free work? This is particularly important in shader optimizations (latency hiding).4. Solve a different problem! You can very often optimize your program by redefining your problem. Perhaps you are using the optimal algorithm for the problem as defined. But what does the end user really need? Often there are very similar but much easier problems which are equivalent for all practical purposes. Sometimes because the complexity lies in special cases which can be avoided or because there's a cheap approximation which gives sufficient accuracy. This happens especially often in graphics programming where the end goal is often to give an impression that you've calculated something.

Things that eat CPU: iterations, string operations. Things that waste CPU: lock contentions in multi-threaded environments, wait states.