Olivia
Online
Davis
@Davis58841504
Joined February 2018
176 Following
87 Followers
78 Posts
SIPs are important, no doubt. But don’t sacrifice life’s precious moments just to maintain every monthly contribution.
A mutual fund SIP is a long-term wealth-building journey, not a 30-day exam you must pass every month no matter what. Missing a contribution occasionally won’t ruin your financial future or goals.
You might miss a month because you’re travelling, spending time with family, celebrating something special, or buying something for yourself or simply dealing with life. That’s okay.
The real power of compounding comes from staying invested and giving your money years to grow not from living with guilt every time you skip a payment.
Build wealth so you can enjoy life, not postpone living in present just to build wealth in future. ❤️
#MutualFunds
I can barely keep up with AI talk in English, curious how much I'd catch in Sanskrit.
Can you imagine a 2-hour conversation between a VC and an AI researcher on AI and LLMs entirely in Sanskrit?
We just recorded one.
Could this be the world's first technical podcast in Sanskrit?
Can't wait to share it. Here is a raw glimpse
The power and copper side of the AI trade is getting hard to ignore.
AI data center deployment will require 212 gigawatts of cumulative incremental capacity between 2026 and 2030, with annual additions scaling from 13.0 gigawatts in 2025 to 55.7 gigawatts by 2030, according to BofA Global Research. Building out each gigawatt of AI capacity costs between $25bn and $45bn, requiring roughly $9.5tn in potential cumulative capital expenditure through 2030. While specialized accelerators and server racks represent roughly 55% to 60% of total site spending, approximately 40% of capital is absorbed by physical facilities outside compute. Power infrastructure accounts for 12% to 18% ($3bn to $8bn per gigawatt), networking demands 8% to 12% ($2bn to $5bn), and specialized liquid cooling systems require 6% to 10% ($2bn to $5bn), shifting capital toward industrial equipment suppliers. Physical resource bottlenecks are tightening as electrification across power grids, data centers, and electric vehicles drives a 2.3% compound annual growth rate in copper demand between 2026 and 2030, matching total global copper demand growth seen across all sectors over the past decade, according to BofA Global Research. Total bottom-up demand of 212 gigawatts implies capital requirements roughly $2.1tn above current consensus hyperscaler budgets of $7.4tn through 2030. If utility interconnects, high-voltage transformers, and local grid capacities fail to deliver power at scale, physical delivery limits will cap server deployments regardless of semiconductor supply, stranding capital in incomplete facility shells and extending project payback periods. #AI #Macro Report date: 2026-10-16 · BofA
@CharlieOchim future me has been leaving me on read lately
Corporate speak can turn a Martha Stewart interview into homework real quick.
I like Martha Stewart so instead of constantly consuming political content I thought I’d listen to her podcast. I thought it’d be a lifestyle podcast but it’s just her interviewing global merchants squatting in America communicating in corporate speak about their “brand”.
These people build nothing. Merchants are the absolute worst class of people on the planet. Today’s episode was her interviewing an Indian named Gupta who owns a website called Quince.
Audiobook chapter markers that start during the last sentence of the previous chapter are so annoying.
Saylor is not exactly the person I'd look to for a neutral read on bitcoin.
Michael Saylor: "We think we're at the beginning of a bull market."
"We've got banking adoption accelerating. We've got tokenized securities and the digital asset space accelerating... all these things are very auspicious for the asset."
Took the long way home and ended up looking at houses again. I do this every time.
I want to see what the clips actually look like after all this.
podcast editors don't know yet they've been replaced by 6 free repos
CLIP LOOM is one html file where three mechanical spiders turn a 2-hour episode into ready-to-post clips. no keys, no backend, no subscription
the idea is simple. a video isn't a picture, it's a thread of sound, and whoever holds the thread cuts the episode wherever they want
most people still hand the raw episode to an editor and wait until friday for clips. i built a loom that runs the whole path in one 15-second loop
how it works:
1. YT-DLP: pulls the episode from one link on thousands of sites, the first runner spider carries the raw reel into the loom
2. WHISPER: turns speech into text in 99 languages right on your machine, no API
3. FASTER-WHISPER: the same text up to 4x faster on less memory
4. WHISPERX: puts a timestamp on every word and runs up to 70x realtime
5. PYANNOTE: splits the thread by voice and knows who spoke at every second
6. AUTO-EDITOR: cuts the silence and pauses, the second runner carries the finished clips to the vault
LINK → TEXT → TIMESTAMPS → VOICES → SILENCE → CLIPS
alone each repo is half the job. one writes the text, one cuts the pauses, one hears the voices. together it's an editing studio that needs neither a human nor a deadline
editing stops being a job and becomes one command in a terminal
bookmark this before someone starts selling it as a subscription
a 0.3% move in the s&p does not need a breaking news alert.
The problem with saving money is that life keeps inventing new things to spend it on.
CIFR looks interesting here. Monthly trend still has me waiting for confirmation before I touch it.
$IREN vs $CIFR have two very different but also similar technical structures. Both have endured substantial corrections amid the broader weakness in AI infrastructure, but an important distinction in where each sits relative to its long-term technical structure.
$IREN hit its macro 1.272 fib at $69 before this correction. A correction of this magnitude isn't particularly unusual, especially following the first test of an important macro extension.
$CIFR fell short of its 1.272 extension at $42 and has now retraced to its 0.887 fib which converges with a significant historical demand zone and the vicinity of its longer-term ascending trendline.
From a purely technical perspective, CIFR's is already testing, and marginally undercutting, an area of support that IREN has yet to revisit on a comparable basis.
For IREN, that equivalent 0.887 retracement sits near $19.40, roughly 30% below current levels, and coincides with the broader trendline structure I've drawn.
That doesn't mean IREN needs to trade there. But it illustrates why I see a more immediately identifiable technical risk/reward setup developing in CIFR.
Sentiment across AI infrastructure has deteriorated considerably, and we're seeing the kind of fear and volatility that historically accompanies important inflection points.
But the indicators I track aren't yet suggesting that the selling is conclusively over. Both of these have lost important monthly EMAs like the 21m EMA which in my experience ends a trend. IREN would need to reclaim $34 by month-end and CIFR $13. Until that happens, one needs to remain open to further weakness in months ahead, even if we experience a sharp relief rally.
And yet, when I consider what these businesses are building, their expanding contractual commitments, and the earnings potential emerging over the next 12-18 months, it's difficult not to be interested at these prices.
It reminds me of April 2025, when several names across the sector lost important technical levels, sentiment collapsed, and what appeared to be a breakdown ultimately developed into one of the more attractive entry opportunities of the cycle. I've followed this sector long enough to recognize that the greatest asymmetries come when the prevailing narrative becomes overwhelmingly negative and price begins discounting outcomes considerably worse than what the underlying fundamentals are showing.
It's not supposed to be comfortable. But I do think we're in an area where the longer-term risk/reward is becoming increasingly compelling, particularly for investors who can look beyond the next several months.
Just be disciplined in your accumulation, position sizing, and be patient.
The amount of arguing over tiny expense ratio differences in beginner investing threads is exhausting.
@FBIDirectorKash @FBI fake account statements are a special kind of betrayal for investors who trusted them
going to be hard for tech to shrug this off if yields keep climbing.
US stocks are mixed around 10:30 to 10:45 AM ET on Oct 15, after September PPI came in at +1.3% MoM versus +0.6% expected.
That’s more than double the forecast.
The S&P 500 and Nasdaq are struggling to find direction, with Big Tech uneven as investors question whether AI spending, debt and valuations have moved too far ahead of actual earnings.
This follows a sharp selloff on Oct 14 after another hot CPI print.
The important part isn’t one ugly inflation report by itself. It’s that both CPI and PPI are now pushing against the idea that inflation is falling cleanly enough for the Fed to keep easing without consequences.
Treasury yields are the first thing I’d watch from here.
If yields keep moving higher, expensive tech gets hit twice. Future earnings are discounted at a higher rate, while companies funding huge AI infrastructure plans face a more expensive cost of capital.
That also makes the dollar important.
A stronger DXY normally adds pressure to dollar priced assets, especially gold and crypto. But gold can still hold up if traders see persistent inflation and fiscal risk as the bigger problem.
$BTC is in a similar spot.
It trades like a liquidity asset when yields jump, but like a scarce macro hedge when confidence in fiat and government debt falls. Which side wins usually depends on whether the move in rates is orderly or turns into a broader risk off event.
Crypto stocks add even more leverage to that trade. Miners and crypto proxies can fall harder than $BTC when yields rise because they also carry equity risk, financing needs and operating costs.
I wouldn’t read too much into every intraday move yet. The market is still open, and the source snapshot doesn’t include reliable live percentage changes for Treasury yields, DXY, gold, $BTC or individual crypto stocks, so I’m not going to invent exact quotes.
But the setup is pretty clear.
Hot CPI started the selloff. Hot PPI confirmed the inflation problem isn’t gone. Now rates decide whether this stays a messy rotation or becomes another broader unwind.
Personally, I’d rather wait for yields and the Nasdaq to settle than buy the first dip just because Big Tech was stronger last month.
YOUR CHASE HOTEL OR AIRLINE CARD MAY NOW GET $120 A YEAR IN FREE DOORDASH
Chase co-brand cards now get a $10 DoorDash discount EVERY MONTH instead of once per quarter.
That's up to $120 a year instead of $40.
The catch is timing
- You only get the monthly version if you activate DashPass with the card for the FIRST time between 9/1/26 and 12/31/27
- Activated before 9/1/26? You're stuck at $10 per quarter
- It's your DashPass activation date that counts, not when you opened the card
So if you have an old Marriott, Hyatt, United, IHG, or Southwest card that you never linked to DashPass, it can still qualify for the monthly $10.
How it works
- Activate DashPass with your eligible Chase co-brand card
- Get 12 months of DashPass free
- Use $10 off one non-restaurant order each month, like grocery, convenience, or retail
- Pay with the same card you used to activate
Important
- Restaurant orders don't count
- Gift cards are excluded
- Unused credit doesn't roll over
- Each DoorDash account can only hold one DashPass at a time
- DashPass auto renews at $9.99 a month after the free year
During the free year, this is basically $10 of free groceries every month.
Set a reminder to cancel before it renews unless you actually use DashPass.
Audiobooks require a task that uses exactly 3% of my brain or I'm gone.
HOW DO YOU LISTEN TO AN AUDIOBOOK WITH ADHD!?
Every THIRTY SECONDS I’m like
“O wasn’t payin attention, rewind”
“Wait crap I got distracted, rewind”
“JESUS I GOT DISTRACTED AGAIN, HERES ANOTHER REWIND”
I prefer audiobooks while doing dishes to listening in bed and losing my place when I fall asleep.
If the $BMBL acquisition had happened literally any other week of the year my net worth would be up 20% today. Had material alpha here both for a sense of timing and the terms of the deal. Unlucky timing for personal reasons.
LL Cool J has been around forever, curious what stories he gets into with Rogan.
Something went wrong.