@coladacodes @AzukiTCG @phygitals u just just buy some bonk
DezXAZ8z7PnrnRJjz3wXBoRgixCa6xjnB7YaB1pPB263
Stefan Mai
5.4K posts
Stefan Mai
@iamnafets
Builder. Older than all dogs. Creating: Hello Interview. Sold 2 startups, failed 1, ready for more. Previously ML eng leader FB, AMZN.
Katılım Ocak 2009
895 Takip Edilen1.1K Takipçiler
We just spent $35k on @AzukiTCG cards, and it's all going to members on @phygitals
If we weren't distributing them, i'd lock them up and sit on it for 5-10 years (nfa but easy play imo)
- great ip
- goated team
- high quality products
- tcgs are booming / here to stay
🫘🥷
Official Azuki TCG@AzukiTCG
Dragon Ascendant tier is sold out! 5 tiers left: tcg.azuki.com
English
@satsdats balanced, keep it up. isn't bonk underpriced at 500m?
DezXAZ8z7PnrnRJjz3wXBoRgixCa6xjnB7YaB1pPB263
English
We are going to be doing a lot of experiments and trying new stuff.
One of the biggest things I see mentioned is liquidity on new launches being thin, and people comparing the liquidity on today's launches vs old coins.
So for traders 0.75% of all post bond volume is going into the liquidity pool, this effectively makes the LPs 4x deeper, and hopefully improves the tokens life cycle.
For deployers, 0.25% of bonding curve volume will go into a rewards pool, which will compound throughout the day.
We can see how this plays out, if it doesn't work we are not going to be afraid to cut it, and rethink.
I see a lot of the suggestions regarding vamps, so maybe something we look at next update.
BONK.fun@bonkfun
As part of our return, we’re excited to announce a new feature aimed at incentivizing higher quality deploys and improve conditions for traders on our platform: Balanced Mode Balanced Mode aligns traders and deployers For traders: 0.75% of all post-bonding volume is compounded into liquidity pools, making LPs ~4x deeper and improving token longevity. For deployers: 0.25% of bonding curve volume is allocated to a rewards pool every 2 hours, compounding throughout the day. Every 24 hours, the full amount is distributed to creators evenly across all successful bonds created within that 24 hour window.
English
The default writing style of ChatGPT makes me nauseous. I’m sick of seeing it in X comments, blog posts, emails, memos and ChatGPT itself.
It’s out of fashion, the flavor has grown stale. I use Claude just to escape it.
Very simple fix for OpenAI: change the secret system prompt to abolish this tiresome tone. Or create a more automatic way for people to give it their own tone.
This feels like an instance where a 10 minute change could make a product 2x better
English
@dwlz No chance, I’d be willing to bet 2 to 1 odds. You might see teams adjusting their usage and posture, but handwritten code is on life support not a possible retreat.
English
After people get a few months of seeing the impact of AI-induced slop in their codebases, and the resulting slowdown this incurs, I predict we're going to see a mass reversion back to "handwritten" code.
TBH I'm feeling it myself some days. I'm actually faster with Cursor tab and just nailing what needs to happen. AI has a terrible habit of spending time on things that just don't matter.
Might be more "work", but in terms of clock time, I'm getting things done more quickly without involving token inference.
Maybe I'm weird. Maybe this is a terrible prediction (as things frequently are with things that change so quickly), but in this case I've been observing my behavior and others for months and I'm seeing a slow steady trickle back to the "old ways".
English
Codex app is surprisingly buggy. Threads that I can never archive. Prompts that never dismiss. Is no one using it?
English
@peer_rich @calcom I think this is a good idea in principle but it biases against your most tenured engineers which can stifle their growth. You need mechanisms to make sure they’re not sunk by their success.
English
one of our internal rules at @calcom:
if you introduce a bug youre usually the best person to fix it
you have the context, the edge cases in your head and the mental model of what you were trying to do
throwing it over the fence to a random bug fixer is usually slower
when a bug reporter says the button doesnt render in safari theyre seeing a symptom
but the original dev remembers why that button was implemented which library was used and which commit touched it last
that context is the difference between a 10 minute fix and a 2 day round trip
context loss is one of the quietest productivity killers in any growing company
the moment you start routing bugs by availability instead of ownership your velocity slows to a halt
you save time on task assignments but lose 10x more in cognitive reload
bugs are inevitable when youre moving fast
but weve found the healthiest patterns are you break it you fix it you see someone struggling you help them fix it and you never outsource responsibility for quality
English
@DynamicWebPaige That's crazy! It's actually frightening the fragility of most of our infrastructure.
English
Our phones calculate where we are by measuring how long it takes a signal from a GPS satellite to reach us -- but the satellites are about ~12,000 miles away, and they transmit with the power of ~50 watts (like a dim lightbulb 😅).
Which means that by the time the signal from a satellite reaches Earth, it is incredibly weak (~one millionth of a billionth of a watt, which is like trying to see a 50-watt bulb from 12,000 miles away). Because the signal is so faint, it is terrifyingly easy to drown out, a cheap jammer bought online for $50 can overpower it locally.
👇And TIL that happens, all the time:
English
extremely proud of our work at Row Zero
we've built an Excel-compatible spreadsheet but 1000x more performant, 1000x bigger data, and cloud-native security
amazing what can be done by a tiny team completely obsessing over a problem for a few years
x.com/GetRowZero/sta…
Row Zero@GetRowZero
AWS is officially a Row Zero customer! They approached us looking for a secure alternative to the mess of CSV & Excel files on laptops, and needed a tool that could handle Amazon's scale AWS chose Row Zero's industry-leading performance and security over 13 alternative tools
English
I maintain there are only 3 kinds of people
Arc benders - Things get better when you look away
Arc setters - Things get done when you look away
Arc riders - Things get worse when you look away
TsvetanK@tzvetan_Ck
Managing 40+ direct reports? Nik Storonsky breaks down his brutally simple framework for evaluating talent Excellent people = self-guided missiles. They pick the goal, press the button, and hit the target themselves. Strong people = you show them the goal, they execute without hand-holding. Average people = weekly iterations needed to reach the goal. Below average = constant iteration, goals still missed. The key? Only hire strong or excellent. No exceptions. Credit: @HarryStebbings's podcast - 20vc.
English
@timneutkens So good. The guess and check of trying to figure out how a module made its way into the route was infuriating! Can't wait to try this.
English
We've been working on something new:
Turbopack / Next.js bundle analyzer.
It's fully aware of Next.js routes and the Turbopack module graph.
You no longer have to guess where a module is loaded.
It shows the full import stack so that you can easily find where the module is used, not just one path but all paths that lead to the module being included in the bundle.
You can also inspect CSS size and other assets, as well as the server bundles.
Note: Output size shown is without compression (gzip/brotli). We're planning to add that.
For example running it on an open source application:
English
@vedantmisra Aw that's so sweet. She's trying to show interest in your work!
English
Mom: how’s work going?
Me: pretty good, we made all the numbers go way up in this table and shipped the best model in the world
Mom: that's nice but what's happening with the "SWE Bench" row
Me: mom can you not right now
English
@peer_rich The insolvency risk you shoulder. It’s a savings account and a credit default swap! So convenient.
English
@headinthebox If your competitor is selling product at a loss, you can win by just buying their product.
English
US tech stocks should go up steeply given adversaries prefer US models over their own.
Anthropic@AnthropicAI
We disrupted a highly sophisticated AI-led espionage campaign. The attack targeted large tech companies, financial institutions, chemical manufacturing companies, and government agencies. We assess with high confidence that the threat actor was a Chinese state-sponsored group.
English
Amazon Echo selling me "premium" white noise when I'm trying to go to sleep is peak dystopia.
English
@ns123abc Depends on where the bottleneck is. The latest generations have substantially improved power efficiency. If you’re capped by watts, not by chips, you’re shedding old hardware quickly.
English
New GPUs do frontier training -> 18 months later do smaller training -> then inference for years after
5-6 year depreciation matches actual economic life, not fraud just how datacenter assets work
Look at the market - even shit-tier compute cards from 5+ years ago held value after Blackwell launched
Depreciation is paper losses for earnings reports. Only matters in a resale but nobody’s selling anyway because there’s a literal shortage lol
Cassandra Unchained@michaeljburry
Understating depreciation by extending useful life of assets artificially boosts earnings -one of the more common frauds of the modern era. Massively ramping capex through purchase of Nvidia chips/servers on a 2-3 yr product cycle should not result in the extension of useful lives of compute equipment. Yet this is exactly what all the hyperscalers have done. By my estimates they will understate depreciation by $176 billion 2026-2028. By 2028, ORCL will overstate earnings 26.9%, META by 20.8%, etc. But it gets worse. More detail coming November 25th. Stay tuned.
English
@jeanqasaur You *needed* good abstractions to move fast previously. Otherwise you're literally typing the same code or similar behaviors in multiple places. That friction is diminished. All that remains is the long-term maintenance costs which are hard to _feel_.
English
To the folks who have experienced production coding standards dropping post-vibe coding, I'm wondering what you see as the main causes.
- Is the volume of code simply overwhelming reviewers?
- Are teams expected to ship at a higher cadence?
- Is it something else?
✨ Jean Yang ✨@jeanqasaur
Curious: have folks seen a drop in coding standards for production since vibe coding came about?
English
@frantzfries Ooh man, can this be a series? Can add my nihilistic correspondence with the Franchise Tax Board.
English
i could deinfluence you on founder life so quickly if you could have witnessed the last two hours I spent with gusto support
English
@MichaelArnaldi What are the largest deployments you've seen of Effect Cluster vs e.g. Temporal/Cadence?
This "reliable at scale" seems pretty fraught to evaluate without substantial deployments for long durations.
English
How do I expect it to evolve in a year or so
Michael Arnaldi@MichaelArnaldi
Landscape of Durable Computing Solutions (yes many solutions missing, similar grouping)
English