Andy Surtees

Been heads-down on this for the last year. Today it's real. If you've ever bounced off Zapier or refused to let an AI agent near your real accounts, this is built for you.

@aakashgupta It's like the original Gloriamundo

Perplexity is a $20 billion company that built zero AI models. Their product sits on top of 19 models made by other companies. Claude for reasoning. Gemini for research. GPT-5.4 for long context. Grok for lightweight tasks. Nano Banana for images. Veo 3.1 for video. You write one prompt. Computer picks the best model combo for the job, spawns sub-agents in parallel, and runs the whole thing in a cloud sandbox while your laptop is closed. 400+ app connectors. Gmail, GitHub, Snowflake, Salesforce, Ahrefs, Shopify. Read and write access. One prompt can scrape your competitors, pull live financials from FactSet, query your data warehouse in plain English, and push a finished report to Google Slides. No API keys. No terminal. The enterprise usage data tells you where this is heading. In January 2025, 90% of enterprise tasks on Perplexity ran on two models. By December, no single model held more than 25% of usage. A new frontier model launched every 17.5 days in 2025. Each one brought different strengths. The era of picking one model is ending. Perplexity built none of the intelligence. They built the routing layer that makes the intelligence usable. Stripe didn't build the banks. Google didn't build the websites. The value is in making complexity disappear. Four of the Mag Seven already use Perplexity's search API in production. Every model provider is now building orchestration in-house. The question is whether the routing layer stays independent or gets absorbed. I wrote the complete guide to using Computer without wasting credits. 6 use cases, the prompt spec that controls cost, honest limitations. aibyaakash.com/p/perplexity-c…

@MapleLeafCap Both of those analyses seem fundamentally dumb to me. Reminds me of the years of "actually, more people use Yahoo than Google". Yeah, but we could see which way that was going

12/ The Writeups If you want to read both pitches side-by-side, here you go: Baseline (Gemini 2.5 Pro, 4o, etc): bidclub.ai/post/a7c3e13e-… Opus4.6: bidclub.ai/post/816f09c0-…

Anthropic claims Claude Opus 4.6 published today is a marked improvement for investment analysis; I put it to the test. I ran it through my automated stock pitch pipeline: 6 AI analysts, 5 rounds of debate, 1 research director. It cost ~$50 (20x my usual). Is it worth it? Here's what I found: 🧵

@Bandcamp Is Ozone AI? Is Melodyne AI? Is RX AI? Is Soothe AI? Is Drumagog AI? Is Autotune AI? Is quantization AI? Is a drum machine AI?

We’re articulating our approach to generative AI so musicians can keep making music and fans can trust the music they find on Bandcamp. blog.bandcamp.com/2026/01/13/kee…

@thisisyourparty They really are creating a new kind of political party. One that involves something like logic, except the opposite.

This is the man Farage and Britain’s leading racists want deported. Alaa was a leader of the struggle against tyranny in Egypt. He risked his life for freedom and democracy, jailed by Egypt’s military dictatorship. The bigots attacking him have no idea what that means.

@thegarrettscott Create a comprehensive, self-maintaining, self-improving knowledge base that enables any human — regardless of starting education level — to achieve functional expertise in any of 100 specified domains within the knowledge base's learning pathways. Validate by taking 1,000

If you have a project that you think is “too hard for AI to do right now” comment it here. I’ll give free Do Anything accounts for the hardest ones so you can see if it works And we’re talking really hard, like: - Go make $10,000 - Start and run a coffee shop in Menlo park etc.

@thegarrettscott Select an unsolved problem from an established list (Clay Mathematics, DARPA Hard Problems, Hilbert-style challenges). Produce a solution or significant advance that survives peer review and independent verification within 36 months. Seriously though, this is great. I'm working.

I am become Oppenheimer, destroyer of reasonable p/e ratios.

@letsplay14me @elonmusk @Vernon3Austin Because the train goes from not where I am to not where I'm going.

@elonmusk @Vernon3Austin Genuine question: What's wrong with cheap, fast, clean, efficient public transport in cities? Most of Europe and parts of Asia do this extremely well. Rubber wheels on asphalt and single-occupant vehicles don't seem to be the future, although I understand the appeal.

Boring Company is dangerously close to being able to provide every city dozens of miles of light rail/BRT type capacity with no public money, limited utilities interruption, and minimal new right-of-way. While offering passenger travel times that are a fraction of other transit.

@bobmarotte @RealTesselator @stepfanie Yeah, this is definitely AI. The (presumably) bot thinks it's Gustavo Panilla, who's another dictator. Other than both being bald, they don't look much alike. Most plausible explanation is that there is a mislabelled dataset somewhere, so it guessed based on the uniform, pose etc

@RealTesselator @stepfanie This is Mussolini, not sure who TF you're talking about.

if you can't name this person stfu about fascism

@sgrami13 @svpino You can't get any serious work done on any laptop, even if it's 17". It's just big enough to trick you into thinking you might be able to. Might as well get a 13" one, do some tinkering on the go, then get it plugged into two 24" monitors at home or the office.

@svpino i need 17 inches and you use 13 ?

This is the best laptop in the world. Period. M4, 32 GB RAM, 1TB SSD, 18 hours of battery life, 13 inches, 2.7 lbs. $1,699 with the student discount. I've used it for 6 months, and I haven't missed a Pro, not even once. This laptop has no competition.

@someRandomDev5 @JuliousNiloy @theo C0omkpare this to Claude and it's clear, but also compare this to GPT3, 3.5, 4 when they were released. Stunning increase in power per dollar in the last few years.

@JuliousNiloy @theo 4.1 Opus is like ten times more expensive and intellectually inferior lol.

Things that stopped being relevant today: - Claude 4 Sonnet - Claude 4.1 Opus - Claude Code - o3-pro - Gemini 2.5 Pro - Gemini 2.0 Flash (2.5 was rough) - Every mini and nano model ever made - More stuff I'm not thinking of

@JuliousNiloy @theo You can always use GLM 4.5 at $0.2/m in, $0.8/m out. It's not quite GPT5 but it's not far off.

hard disagree here even if GPT-5 is the best model atm, most people can't afford it all the time due to the high api cost/rate limits people will naturally fallback to lower models as needed just because the best model launched does not make all other models irrelevant instantly

In 2010, the average house in London cost 2 million Bitcoin. Today it costs 6 bitcoin.

GPT5 Isn't available for most ChatGPT users in the UK yet. But it is available on GloriaMundo - at the standard API price, with no markup.

Grok 4 is now live on GloriaMundo!

This is a lot of fun - create a free bot in seconds, style it, share it, embed it on a website gloriamundo.com/freebots/manage

🌀 Welcome to the AI Multiverse! Lost? Explore infinite AIs and unleash your creativity with GloriaMundo—subscription-free! Your sanity thanks you. Step into coherence → [gloriamundo.com]

👁️ Is the singularity watching YOU? AI models multiplying, subscriptions growing... reality bending? GloriaMundo is your refuge—339+ AIs, one app, no subscriptions. Because sanity is priceless, even if reality is questionable. Question everything → [gloriamundo.com]

🔄 Feeling lost in the endless maze of AI subscriptions? Step away from confusion and reclaim clarity with GloriaMundo—every AI you need under one roof. No subscriptions, no headaches. Just sanity restored. Explore the simplicity → [gloriamundo.com]