Advait Shroff

@premdebrief It has to be: 1 of Tchouameni / Anderson / Tonali (locked in as the PL and UCL starter next to Mainoo), M Fernandes + Baleba (rotation, depth, and cup game starters), Hall or Maxi Araujo (latter may honestly be better), LW (Ndiaye/Godts/El Mala/Diomande) ST (Watkins/Welbeck)

Assuming Tchouameni isn’t available Is there a better window than: M Fernandes Baleba Camara Hall Diomande/Barcola Think this can be done for between 250-300 which is what United will spend I imagine (optimistic)

thank you @googlemaps

@ItzSuds Nikesh is insurmountable

If Anjney Midha actually raises $10b this year, Nikesh Arora’s reign as the most auramaxxed Indian in tech will be under siege for the first time in like a decade (maybe Satya Nadella in ‘24-‘25) Generational one man run right now, turbo’ing AUM + hitting on every single deal…

Interesting paradigm shift to think about: attention vs tokens. In recent times, monetizing attention has been the primary focus. That’s what all social media, and ads boil down to. In this economy, time was the primary currency. But in the new economy, tokens are the primary currency. And you are either a provider of tokens or a consumer of tokens. If you are a large provider of tokens (the labs, and apps like Cursor or Perplexity), the metric that will matter instead of attention share, or market share, is token share. Token Share: what % of all tokens consumed in the economy were tokens provided by you, relative to other token providers? This is because you don’t need active attention to be consuming tokens anymore. You could have long-running, scheduled workflows etc consuming tokens in the background while your active attention is elsewhere for extended periods of time. And that fundamentally shifts the paradigm on which currency matters going forward.

As of tonight, there are two frontier AI labs whose models are approved for use in classified military systems. One has been declared a Huawei-level threat to national security for insisting that its models not be used for mass domestic surveillance and fully autonomous weapons. The other just signed a deal that allows its models to be used for everything except mass domestic surveillance and fully autonomous weapons. I am truly not taking sides here, I'm just very confused.

The DOW got so mad at Anthropic for insisting on carve-outs for mass domestic surveillance and autonomous weapons that they declared it a supply chain risk and struck a deal with OpenAI that includes...the exact same carve-outs.

I use WisprFlow a lot - 100K+ words since I started using it ~2 months ago. I’m a huge fan, but it is definitely not as perfect as it is being made out to be here. Still gets simple stuff like “will” vs “we’ll” wrong with no learning or adaptation. I often ask it to write something in brackets expecting “(xyz)” but it will literally write “brackets xyz” and it irks me to death - filed bugs for both!

We offered 5 people a Porsche 911 GT3 RS if they could get @WisprFlow to make a mistake It's the fastest and most accurate AI voice dictation app that's 3x more accurate than ChatGPT, Claude, or Siri. Today, we’re finally launching on Android. Download now: play.google.com/store/apps/det… As a part of the launch, we’re giving away 6 months of Wispr Flow Pro for free. Like, retweet and comment ‘Wispr Flow’ to get it. Enjoy. — Written with Wispr Flow

@Shealacey10 Keep your head up young king, that hunger is what sets you apart. lots more to come 🫡 🔴

I want to apologise to all of my teammates staff and the fans 😔 I let everybody down tonight I shouldn’t let my emotions get the better of me I will do everything I can to make this right 🙏🏼

@snareyansh @chinmayjindal_ @GroqInc @nvidia @scale_AI Opened + just DM-ed you!

@advaituncombed @chinmayjindal_ @GroqInc @nvidia @scale_AI hey advait! there are some roles im interested in, but your dms are not open

Hey everyone, I wanted to share an update regarding my offer at @GroqInc / @nvidia. I was set to join Groq in Summer 2026 to work on distributed systems and software infrastructure that support their next-generation inference technology. However, following the recent $20B agreement between @nvidia and @GroqInc, my offer was unfortunately rescinded. While the timing is disappointing, being selected to work on low-level systems supporting large-scale inference was a meaningful validation of the work I’ve been doing in systems and ML infrastructure. I’m grateful for the opportunity and for the engineers and recruiters I met throughout the process. That being said, I am now actively recruiting for Summer 2026 roles centred around ML systems, training infra, or foundational models. If your team or someone you know is building in this space, I’d love to connect and talk about my past work. My DMs are open.

This is some of the craziest shit ever, if true. I guess it’s hard to be surprised when all corporations are forced to be profit maximizing entities out of fiduciary responsibility but Jesus Christ still. Have to imagine that there are several real benefits to having actual driver benefit programs and just treating them with even the most basic decency would lead to higher retention than all of the elaborate schemes. Hoping for a WSJ exposé in a few months here

holy fucking shit

@businessbarista Claude

I want to start a community dedicated to Claude Code. It’s become the gateway drug to coding and experiencing the power of AI for tons of people. This will be a space for people to share killer use cases, agentic workflows, proven prompts, and connect with other CC obsessives. Comment “Claude” if you want to join.

But honestly though why does he suck so much Congrats to @priyashah_ and team on shipping this joy of an experience

Can AI handle the kind of reasoning professionals rely on daily? Our latest benchmark, PRBench, puts models to the test with over 1,000 expert-authored tasks in finance and law. Even the strongest models scored below 40% on the hardest tasks, highlighting the gap between potential and practice.

Only good things can happen when an epic team joins another epic team! So excited for what this means for the future of AI Native computing 🚀 and congrats to the whole team at @SoftwareAppsInc 🐐

@pitdesi @ScottAdamsSays This is excellent advice. Sounds kinda familiar though…

Rules of good writing by @ScottAdamsSays

LINA FROM THE TOP ROPE

@utdjohnyy Onana

What’s stopping this team from winning the League ?

@pitdesi I think there’s an interesting anecdote about how Wag’s pitch deck for Masa had a picture of Masa’s dog whom he loves, which played to his emotions. They were writing checks for fun with little diligence in those days but this story makes Wag an especially interesting one

Wag is a perfect cautionary tale of raising too much money. They had a direct competitor in Rover that never raised a mega-round… and ended up getting acquired for $2.3 billion. Wag raised $300 million from SoftBank in 2018, back when SoftBank was firehosing cash into every “Uber for X” startup: Getaround, Alto, Reef Kitchens, Zume Pizza, Tier, Wag, DoorDash, Ola, Grab, Didi, Convoy. It did not end well for the majority of them, and is absolutely wild to think how much money was lit on fire there. When you raise that much money, you can skip the hard part: finding product-market fit. You just scale whatever you’ve got, whether or not it works. Dog walking was probably never a good standalone business. It's low margin, highly recurring, and has little lock-in. My friends that used Wag all went off-platform and venmo'd their walker directly. Rover focused on overnight boarding and house-sitting. Bigger bookings, higher margins, and more need for platform protections like insurance and support... a wayyyy better business model I feel like that $300M check fairly early in the companies life let them paper over bad retention, high CAC, and leakage with an insane amount of marketing spend.... but ultimately Softbank wasn't there for another round, they went public via SPAC and never made any profits... the music stopped and they were left with a shitty business. Meanwhile, Rover raised gradually and had time to test, iterate, and find what worked before pouring gas on it. IMO slow money made them better. They discontinued on-demand dog walking shortly after they started it because it had poor unit economics... If Wag hadn’t raised $300 million, they might’ve figured that out too? Pour one out for Wag.

@UTDTrey @ChelseaFC @ManCity @Arsenal @LFC @SpursOfficial @FCBarcelona @realmadrid @Atleti @Saudi @rihanna @Inter @juventusfc @sscnapoli @Atalanta_BC @FCBayern @BVB @bayer04_en @PSG_English @AS_Monaco @InterMiamiCF @Alhilal_FC @AlNassrFC_EN @rihanna is crazy 😭

🚨@ChelseaFC @ManCity @Arsenal @LFC @SpursOfficial @FCBarcelona @realmadrid @Atleti @Saudi @rihanna @Inter @juventusfc @sscnapoli @Atalanta_BC @FCBayern @BVB @bayer04_en @PSG_English @AS_Monaco @InterMiamiCF @Alhilal_FC @AlNassrFC_EN SOMEBODY HELP US PLEASE