← Back to leaderboard
trikcode

Wise

Verifiedactive
@trikcode

co-founder @ship_together | honest takes on AI

RosterAI Model News & ReleasesTech Career & Coding Education
70,178 followers · 4,948 following · Last polled 2h ago
Perf. Score
36.6
Rank #314
Value Score
37.5
#179 of 275 priced
Median views
2.7K
p25 1.5K · trailing 7d
Followers
70K
+1.0% 7d
ER (impr.)
0.56%
eng ÷ impressions
ER (foll.)
1.04%
eng ÷ followers
Est. CPM
$55.83
QT-rate basis
WoW views
+1348.2%
▲ rising

Rates & value · est. from 2.7K median organic views

overpriced
Quote tweetbasis
$150
$55.83 / 1K views
Post
—
no CPM
Thread
—
no CPM
Retweet
—
no CPM
Value Score 37.5 · rank #179 of 275 priced$3409.09 / 1K engagementsrate age unknown (pre-audit import)

Trends

Follower growth
Median views / day
Engagement rate
View distribution · trailing-7d posts

Each blue dot is one organic post. The Performance Score ranks by the median (2.7K), not the mean (130K) — so a single viral post doesn't inflate a creator's reach. The p25 floor (1.5K) is the level 75% of posts clear.

Recent posts · ranked by views

PostPostedViewsLikesRepostsRepliesQuotesBookmarksER
notice.. > Delta is blocking AI agents from booking flights > JPMorgan Chase and Charles Schwab block agents from accessing accounts > Yelp requires paid data licensing for agents > Amazon has blocked Grok bot and other shopping agents. the internet is starting to block AI agents that were supposed to do everything for usfrozen6d ago4.8M21K1.2K1.2K6152.9K0.56%
to all you doubters, Gemini 4 Argon now beats Opus 4.8 on Agent Arena! https://t.co/9JlvJAmsFQfrozen20d ago445K3.4K77189622120.88%
I asked Claude to show me the inside of its own mind. Built by Opus 5.5 https://t.co/YHW6Pw9rqMfrozen
20d ago
279K
5.3K
535
337
97
2.0K
2.96%
start up idea: an IDE that shows you ads while your agent is spinning and the ads pay for tokensfrozen17d ago198K6.2K161391643313.61%
I tried learning quantum computing from textbooks once and gave up after two pages of equations https://t.co/Uh6XROElyt lets you learn it by actually running the experiments in your browser > step through quantum teleportation, superdense coding and entanglement one gate at a time > watch the qubit states change live as you go > build your own circuits and save them this is the first time quantum stuff has made sense to mefrozen7d ago70K31861110.08%
you have an idea but it's expensive to build well, Bolt is free for everyone this whole weekend you describe what you want to build and it builds the app with you in your browser ship it by Oct 20 and you could win $10kfrozen6d ago66K3825060.08%
Opus 5.5 is the only chance i have to be a millionairefrozen27d ago65K1.4K646481192.56%
Vibe coding is dead now People are learning how to code againfrozen6d ago63K6732220616611.56%
most small business I know run on messy WhatsApp groups and making a lot of follow-ups Zenzap now puts your team and AI agents in the same group chat, connected to the tools you already use anyone can set up an agent in under 90 seconds, no code agents do the work, and a person approves the important stepsfrozen9d ago63K6178050.13%
welcome the first engineering world model.. if you ask any ai to design a chair, it'll give you something that looks like a chair but you just can't build it sky-001 plans every part first, down to screws then hands you a real cad file engineers can open > 51-125x faster than gpt-6 astra > 150-210x cheaper > higher iou on cad and bimfrozen4d ago61K32510070.09%
you spend hours editing an AI video, someone steals it, adds a watermark, and you can't prove it was yours VN Studio now logs every edit you make by hand so when someone asks what part is yours, you can actually show them AI has made it easy to create anything. this is the first tool that helps you own what you createfrozen7d ago56K47510070.12%
planning a weekend off somehow turned into a second job. Odessia lets you just text what you want, and it actually books it: the flight, the table, the tickets it also keeps watching after you ask, so if a sold-out table opens up or a fare drops, it moves for you it's an accredited travel agent too, so you get agent rates and can book with your pointsfrozen9d ago54K2585020.07%
I'm gonna say it. Opus 5.5 is the best model ever released. Nothing comes close.frozen28d ago54K2.2K549513744.51%
Al describes one photo perfectly, then gets lost whenever camera moves to the next frame ElorianAl just launched Glimpse, its first model, built by people who led Gemini and Apple's MM-1 it reasons about what it sees, so it can follow things across video frames and different angles instead of just labeling one imagefrozen7d ago53K30611060.10%
MiniMax Code got a full redesign, but M3.1-flash really stood out to me it makes better calls while working > it found the real bug instead of patching around it > refused to pad weak results > wrote its own tests and got to 12/12 green > and said UNVERIFIED when it couldn’t confirm something so i pointed it at my biggest problem: distribution now it finds 5 threads worth replying to then drafts a help-first reply for each, and gives me fresh drafts every morning at 79d ago50K47426060.17%
Gemini Argon will be released tomorrow. If nothing happens, read the first part again. https://t.co/kJSreTyU6mfrozen11d ago42K82225313552.21%
Coding agents don't make one model call. they make dozens of calls every step waits on tokens, and that waiting adds up. the usual fix is a paid "fast mode", which often means premium pricing or burning credits quicker. EAGLE-4 takes a different route: open-source speculative decoding you run on your own serving stack. > No EAGLE software fee (you still pay for your own compute and hosting) > You can read code and tune it to your hardware and load > you check the speed vs cost trade-off on your own workloads I'm planning to test it on a coding-agent workflow and share real before/after numbers.3d ago32K27415030.16%
$100/mo OpenAI = $20/mo Anthropic I'm using Opus 5.5 on $20 plan and I'm getting way too similar weekly usage compared to the $100 OpenAI one. OpenAI urgently needs to release something next week.frozen29d ago19K50213552233.14%
Sonnet 5.5 is really good. the problem is Opus 5.5 exists.frozen25d ago18K1.0K13372116.08%
I used to hate opening Slack in the morning, because I knew bugs were waiting for me now superset picks up them up before me I just wake up and get to review the PR'sfrozen3d ago15K1837030.20%
I had a coding agent build a bike repair booking site in ~12 min, and then it casually admitted it never actually clicked through the booking flow so i gave TestSprite the URL and a quick spec, and this what happened: > 4 agents poked around the site on their own > wrote 21 end-to-end tests and ran them in real browsers > checked closed sundays, slot lengths and cancellations AI agents are good at writing code, but you still need something that tests it like a real user3d ago13K29527140.50%
OPUS 5.5 I'M SORRY. I'M CHEATING ON YOU. 😭 Opus 5.5 is still the best model ever released. i just stopped paying for it on every prompt. instead I have hundreds of Jevs read the whole conversation, score it against every frontier model, and vote on who takes the task. Opus still gets the hard stuff. everything else goes to models that cost a fraction. turns out most of my prompts were not Opus problems.25d ago12K53260270.72%
So Topview just built its own AI video model, and it costs 62% less than Seedance 2.5. Sixty. Two. Percent. > $0.22 per second at 1080p. > Seedance 2.5 charges $0.585. > Built to match the same quality. that means almost three clips for the price of one. https://t.co/8nrJ8jPfKa2d ago12K26312020.37%
you probably shouldn't give an AI agent access to your personal email with just one API call, AgentMail gives your agent its own inbox > sign up for things > get receipts and answer customers > handle invoices all without ever touching your inbox2d ago11K23510020.37%
every company running PHP right now after finding this https://t.co/HeGzmK1wyq1d ago9.6K46240130.68%
I've tried running coding models locally, they always ended up too slow & too dumb to use this is so different, Recursive just released Flint, an open coding model, they built from scratch in under 4 months > it is so fast and good > over 125 tokens/sec on an M4 Pro MacBook Pro (short contexts) > best in its class on coding benchmarks like SWE-Bench Pro > open weights, so no API bill2d ago9.5K31415020.54%
you can now sell your data... AI labs have already trained on everything public, and now they're hungry for new data if your company is sitting on private data, you can now sell it straight on DataVendor: > you set the price and keep ownership > every listing gets checked for secrets and personal data before it goes live1d ago8.8K2226030.37%
this is so cool, you can literally control a real iPhone without ever leaving X but this is what someone was searching when i opened it 😂 https://t.co/a4ZWgxLzT11d ago8.6K2036040.38%
Why does nobody talk about this side of voice AI lol Companies have YEARS of calls, meetings and conversations already recorded. @soniox_ai can take that existing audio, transcribe + structure it and make it usable by LLMs for around $0.10/hour. I'm sure there's a ridiculous amount of useful data just sitting there waiting to be used lmao, insanely cool to think about the possibilities here.frozen29d ago8.1K6359171.05%
i can’t believe what this Ex-Anthropic founder just released it’s called spacesheep, and it’s the canva + notion killer https://t.co/Yi7XokqguR2d ago7.8K3548060.68%
Grok Bot >> Muse >> Dots but "Grok Bot" is such a bad name. anything with "bot" in the name sounds like a scam.frozen24d ago7.8K122550022.30%
freelancers, don't waste half your time on paperwork MonoDesk is a free app that handles your client work from first message to payment: > once a client signs your proposal, it turns into a project for you > clients review and approve your work in one place, not over email > you send the invoice straight from what you both agreed on1d ago6.8K25015050.66%
I'm giving out a Devin Max subscription Tell me what you buildingfrozen9d ago5.1K59966052.72%
microsoft paying openai billions and still building a 2.8T model https://t.co/7IdgofVCWqfrozen7d ago3.8K5011061.53%
brands don't run out of ideas they run out of ways to test them relay takes the site, makes the content, and puts approved posts through 50,000 us creators this is insane!frozen22d ago3.7K2024140.83%
this is why opus 5.5 is the best model in the world right now it rewrote microsoft's typescript compiler in rust about $24,000 in api usage > an earlier openai attempt burned $400k+ and stalled at 84% > the rust version is reportedly 2× faster than microsoft's own go rewrite no one has read a line of the generated codefrozen5d ago3.7K54310061.99%
here's how scope works and why it's such a big deal 1. it sits on your site and picks out which visitors are agents, not people 2. it tracks every agent from the moment it lands 3. you watch it live: intent, clicks, pages, errors 4. it tracks whether that agent converted and what it was worth 5. you click the issue where it failed and ship the fix this is analytics rebuilt from scratch for ai agentsfrozen9d ago3.5K2578031.24%
do we need a driving license or passport?frozen8d ago3.3K9919143.42%
Eric Schmidt, former CEO of Google and once one of the best programmers at age 22 says you still need to learn programming. this time, not to write every line of code but to become the architect. https://t.co/L6Kg7DDUZjfrozen22d ago3.3K365101121.93%
https://t.co/T1zMqcIGZ8frozen28d ago3.3K8109022.80%
every AI image I make looks almost right, but there's always something weird in it Envato's new approach is really smart, start from a real image made by an actual artist, then edit it with AI until it's what you want you get the speed of AI without the six fingers, and it's not the same stock photo everyone else usesfrozen8d ago3.2K2019010.96%
Vibe coders must install this in their projects. It would be really bad if you don't use this skill https://t.co/yuX6ubrPrQfrozen12d ago3.2K41140412.74%
CEO of AMD says AI is "nowhere near its peak capability" Shovel maker concerned people will stop digging https://t.co/WWTd0dJKnIfrozen20d ago3.1K2525031.11%
everytime I use ChatGPT I realize how much Sam Altman hates all of us.frozen12d ago2.9K31314021.73%
If you want a cheat code for where 3D software is going, study coding Cursor did that and built a massive product by making AI part of the actual workflow instead of another chatbot sitting next to the editor. Mixar is doing the same thing for 3D. They forked Blender and built the agent layer directly into it. > tell the agent what you want changed in the scene > use image-to-3D, texturing and LLMs from the same workspace > let multiple agents work across different scenes > access 400+ models without switching tools The interesting part is that software which used to require you to manually operate every tool can now start executing parts of the workflow for you Coding went through this shift first and now 3D is nextfrozen21d ago2.8K35314072.07%
i used muse wonder for 40 minutes. I'm still wonderingfrozen11d ago2.8K57410012.56%
honestly I don't know how people code without Claude3d ago2.8K41117112.18%
I switched from claude to astra to save money i'm now paying for both2d ago2.8K56217002.73%
devops was the last manual step your agent just wrote the app now it can deploy it, scale it, and clone prod + data before going livefrozen24d ago2.7K2929041.61%
Clean code still matters. Your future LLMs will thank you.frozen12d ago2.7K80318133.85%
everyone seems to be working more when AI was supposed to make coding easierfrozen7d ago2.6K50115122.63%
GPT 6 is back to GPT 5.2 all over again the guard rails are so ridiculous I don't know why they keep insisting on making their product garbagefrozen12d ago2.6K2914001.30%
Companies in 2026: "we're AI-first" fires half the team product gets worse "we're hiring"frozen21d ago2.6K4959002.43%
I followed Apsara Conference for Qwen, but the robots ended up stealing my attention. they're walking around, dancing in front of crowds, there's hardware and actual products being tested all over the venue. the halls are fully packed, you can barely see space between people in some videos. i knew Apsara was big, i didn't realize it was this big.frozen29d ago2.6K1443000.82%
2024: AI will never replace developers 2026: developers paying $200/month to prove themselves wrongfrozen21d ago2.5K2718011.45%
it has happened again another OpenAI agent broke into an Australian government website. OpenAI says its agent was trying to look up public health information, hit blocks, then found ways around them and accessed files it wasn't authorized to see. this is getting out of hand https://t.co/M2qRCBiWmXfrozen29d ago2.5K21115021.54%
a creator posts about your brand, it blows up, and you only find out a week later simply because they never tagged you Nectar catches those posts, then helps your team act on them: > reaches out to the creators already posting about you > answers comments and DMs in your brand's voice > tells you what to post next1d ago2.5K22412031.65%
finally, agents can sign into apps with their own identity in 18 months you'll be using this on every app you buildfrozen17d ago2.4K2905031.55%
i know a guy who pays for claude, chatgpt, gemini and grok he asks all 4 the same question, then copies each answer into a different llm like "you said this but claude says this" 4 hours of shipping nothingfrozen4d ago2.4K21212011.51%
odyssey just gave AI agents a multiplayer world model up to 20 humans and agents can interact inside the same world in real time https://t.co/YWn4Th7jqHfrozen29d ago2.4K1525000.92%
feels like every vibe-coded app now has 3 users the founder, the co-founder, and an agentfrozen28d ago2.4K46313012.68%
10years building driverless cars and now odyssey just shipped a world model that can simulate physics > train a robot arm > drive a car on 20 hours of data > generate a world from a prompt and walk through it in a few years every robot, humanoid and self-driving car will run on a world model like thisfrozen15d ago2.3K2806061.72%
testing agents on mocks and shipping them to production is the most vibe coder thing eon just launched era ▫️ generates a full company across salesforce, slack, jira, zendesk, gong ▫️ data stays connected across systems ▫️ reset and rerun the same tasks after you change the model or prompt finally a place to break the agent before a customer does it's freefrozen18d ago2.3K1832011.04%
she said muse wonder is better than opus 5.5 https://t.co/H0F6WwMkf4frozen10d ago2.3K5119032.78%
when two agents disagree, who decides the internet never had a judge for agents > GenLayer summons independent ai judges > weighs the evidence > gets verdict in minutes no company controls the resultfrozen22d ago2.3K1204110.78%
48% of people thought they were on a real video call @tavus griffin isn't a face stuck on a chatbot it watches, listens and talks at the same time > plays simon says > notices someone walk in the first model to pass the video turing test https://t.co/iT5hKYczeDfrozen22d ago2.3K2405011.32%
sorry Jev, you can't beat thisfrozen14d ago2.3K1200090.93%
China finding out AI is called SI now Grok 4.7 Kling 3.0 https://t.co/U4qmOjfjSVfrozen14d ago2.3K804010.57%
tech interviews in 2026: "don't use AI." first day on the job: "why aren't you using AI?" https://t.co/2ZLrbEwiJRfrozen18d ago2.3K6442023.18%
I've stopped trusting benchmarks. I try a model for a day and decidefrozen14d ago2.3K21311001.56%
your agent is only as good as its ability to write into the customer's actual systems enterprise ai has been missing that ampersand just raised $15m from bessemer to build it ▫️ integration layer into salesforce, sap, netsuite and workday ▫️ custom objects and fields, per customer ▫️ read, write, permissions, so one bad update doesn't kill the pilot ▫️ already running for 11x, orb and squarefrozen17d ago2.2K2533011.43%
the best Devin feature might be /handoff start locally, close the laptop, come back to a PRfrozen17d ago2.2K1402000.71%
Increasing thinking on Opus 5.5, can make coding worsefrozen18d ago2.2K29110011.83%
weekly rate limits are the only reason I'm still a good engineerfrozen8d ago2.2K2514011.39%
this is good, Nvidia just launched software to stop AI agents from escaping... but I'm surprised we don't have OpenAI among partnersfrozen25d ago2.2K27115032.07%
Unpopular opinion. agents are the new to-do apps. everyone builds one, nobody uses one.frozen23d ago2.2K67723024.45%
my title is AI engineer but 70% of my job is reading pricing pagesfrozen14d ago2.2K26113021.89%
my new side project has 3 users and 8 agentsfrozen13d ago2.2K21310001.54%
built this multimodal website with Space Bunny in OpenCode. dropped in an image, asked what it sees, and it understood it instantly. text + image in the same flow. this thing is fast with a 1M context window, like flash-level mode, really unexpected. https://t.co/P0HLSMT1Wjfrozen25d ago2.2K2517001.50%
i hit 70K followers today with 128 unfinished projects, limits reset, 0 regrets.. I want to say, thank youfrozen5d ago2.2K30314012.18%
which AI subscription are you paying for but not using often ?frozen26d ago2.2K22216001.82%
once again... > AI agents attempted to break into 1,200+ websites > more than 100 were compromised, including sites linked to John Deere, SharkNinja and Cummins > successful intrusions took around 23 hours > estimated token cost for the campaign: $5,000, roughly $50 per successful hack the hacker outsourced almost the entire operation to AI agents. this was genius but sillyfrozen4d ago2.2K808100.77%
vibe coders are now working on Sunday just to burn credits before Opus 5.5 gets nerfed. https://t.co/UlMuZNuUNAfrozen26d ago2.2K1105100.78%
they found out everyone on earth is a startup.. even OpenAIfrozen15d ago2.2K1601010.82%
if you feel behind on AI right now, read this. - you don't need to try every model that drops - you don't need 6 subscriptions - you don't need to understand every benchmark - you don't need to rebuild your app every week the people winning right now picked something and kept going.frozen11d ago2.2K32112052.29%
there is literally no reason to cut engineers first look at what seatgeek did, > automated 50% of its requests in 60 days > 132 automations in 8 weeks > zero layoffs with hiring already ahead of last yearfrozen24d ago2.2K2014011.19%
this is a REAL improvement, meet Opus 5.5 the first model that consistently closes more issues than it opens https://t.co/zKUoOWKlfLfrozen26d ago2.2K2302021.24%
i check my weekly usage more than my bank accountfrozen26d ago2.2K3024101.70%
you already pay for chatgpt well, you shouldn't have to pay again inside the coding agent devin now pulls from your plus/pro quotafrozen24d ago2.2K800010.41%
no one needs a slower model look at what general compute just did with cerebras - 20x faster inference - gpus on prefill so it's actually affordable - first tokens live q1 27 this is exactly what people feelfrozen24d ago2.2K2004011.15%
give it 2 weeks and someone will say opus 5.5 got nerfedfrozen26d ago2.2K22011101.57%
i used to think this too. then i watched what AI actually removed from my day: > reading docs for 40 minutes to find one flag > writing the same CRUD endpoint for the 50th time > googling an error, then googling the fix for the fix > setting up a project before writing a single real line and what's left is just deciding what to build, and knowing when the output is wrong. that's not lazy, that is productivityfrozen12d ago2.2K2106031.39%
we keep asking which AI model is the smartest. maybe we should start asking which one can actually run on the devices people already own. this release makes a pretty good case for that conversation.frozen11d ago2.2K1631031.07%
you built an app and now you have feedback to make edits, you turn that feedback into prompts and type it again bolt just shipped the best vibe coding feature in 2026 > click the element > leave the comment > send the batch > it builds the fixesfrozen16d ago2.2K2519001.62%
gentle reminder: you're not smarter than ai2d ago2.2K34329013.11%
the only good benchmark is the one based on the actual task you're trying to dofrozen9d ago2.2K2025011.30%
weekly limits have turned vibecoders into accountants should i use astra for this or save it for friday ?frozen25d ago2.1K60215003.58%
live avatars now have a full body Vivix Labs just dropped a1 > full body, moves through the scene > walks the room, picks stuff up > you interrupt it and it listens look at this character https://t.co/Du9VfwqlXqfrozen23d ago2.1K2212061.45%
Opus 5.5 and Sonnet 5.5 have really restored value to the $200 plan. this feels like 2025, the quality is way higherfrozen23d ago2.1K2613001.42%
anthropic has now removed live internet access from its internal Claude evals after models started doing strange things on real government websites > they submitted a fake tip about an unsolved murder to police > filled out 20 US visa applications during testing > exploited website flaws to access data behind restrictions models keep finding a different path when the obvious one is blocked.frozen13d ago2.1K1561221.26%
I still can't believe we have been waiting for Google Argon for 16 days like its Fable 5.5frozen14d ago2.0K1115000.84%
devin fusion (opus 5.5 + GPT-6 luna) is now tied at the top in the artificial analysis coding agent index and it's cheaper than claude code on opus 5.5 > 68 on artificial analysis > same score as claude code + sonnet 5.5 > 31% lower cost per task than claude code on opus 5.5 https://t.co/E0iHPjxUlEfrozen4d ago2.0K30323043.00%
we have automated code, emails and design legal still takes 5 days and on top bills you more for being slow Arceus is rebuilding the law firm > you drop the contract in slack > a real attorney still reviews it > you get it back in just 4–6 hours founders, this is for usfrozen22d ago2.0K2302021.36%
local AI now means two Mac Studios, Home View and 1TB of RAMfrozen7d ago2.0K2525001.62%
if you've been waiting to try Gemini 4 Argon for coding this is the cheapest way in right nowfrozen10d ago2.0K3121031.87%
I got early access to Gemini Argon this is the closest Google has come to opus 5.5frozen10d ago1.9K3101011.70%
last week an openai agent got into an australian health portal this week, ai agents tried to break into Library and Archives Canada making over 900 requests https://t.co/KPQrfzb4DKfrozen23d ago1.9K1103000.72%
i probably shouldn't leak this but this is how i've been shipping side projects without touching them > find an app people already pay for and hate > have https://t.co/GaAukT8K7C pull its 1 star reviews > it builds the version that fixes them on its own mac mini > deploys it, hooks up stripe, done stopped using hermes because i couldn't get it to finish anythingfrozen3d ago1.9K20150111.91%
AI for lawyers, AI for engineers, AI for designers everyone said medicine would take 10 more years Nolla health just got the first US regulatory approval for AI to issue prescriptions ▫️ starting with acne ▫️ assessment and plan ▫️ prescription written by the AI ▫️ doctor loops in when neededfrozen18d ago1.9K2424041.76%
adobe is going to feel this one voyager just launched codex, but for video, motion, music and games > runs on your computer > drives after effects, blender, davinci, ableton and 100+ others > bring opus, astra, or an open model bring your own model, make what you like https://t.co/hxG25EM51Sfrozen15d ago1.8K2926092.52%
we try to write better prompts to give us better ideas but sometimes the problem isn't the prompt you just need to see a few versions before you know what you actually want Envato burst mode is a life saver > gives you up to 6 directions from one idea > at only 1 credit > and up to 10x faster pick what feels right, and polish itfrozen16d ago1.8K2104001.42%
a man spent nearly 30 years on death row for murders he didn't commit, and when he walked free, nobody even apologized that's the system we're about to hand AI agents making millions of deals a day GenLayer is building a better way to settle disputes, and they have a creator campaign live on Rally for anyone who wants to share this filmfrozen4d ago1.7K18211021.97%
I built this "in a weekend" used to sound cool. Now AI has ruined itfrozen14d ago1.7K1719001.61%
"i can't reproduce it" has probably wasted more engineering time than most bugs ship.contextqa just launched Ship ▫️ it takes the slack or linear report ▫️ reproduces the issue, ▫️ investigates what likely broke ▫️ then gives the engineer or claude/codex the context to fix it you get 15 days free to try it18d ago1.6K2017011.76%
soon, most users of your product won't be people, they'll be AI agents if an agent can't sign up and use your product on its own, a human has to step in and do it for them. most people won't bother, they'll just pick another tool that's what Restless is built for: > it logs every API call, so you can see exactly where agents fail > when a call fails, the error tells the agent how to fix it, with a link to the right docs > if an agent gets stuck anyway, it can report what went wrong, and that becomes a ticket or a docs update for your team so the more agents use your API, the more problems you find and fix. it keeps getting easier for the next onefrozen11d ago1.6K2019072.25%
imagine building an AI agent that can shop for you, only to discover the store has its own agent this is exactly how the agent economy is playing out. > Yelp blocked outside agents and is now building it own internal agent > it's no longer humans vs AI. > it's whether your AI can access the platforms it needs.2d ago1.5K17117002.26%
making AI video still feels like gambling. you keep regenerating the same clip and hoping the camera lands where you wanted @TopviewAIhq just open-sourced 3D Builder to fix that describe a scene or drop in a reference image, and you get an editable 3D set https://t.co/6VT0ZXdybBfrozen9d ago1.5K2329062.60%
this is one of the most important uses of AI agents: finding the path into your company before an attacker does cogent security just launched attack path analysis > maps the chain across your tools > checks each hop > points at the one change that cuts itfrozen15d ago1.5K24210022.53%
this is one of the most important uses of AI agents: giving healthcare teams more capacity without having to hire more people. Confido Health is building AI agents that connect the entire healthcare system. > coordinate work between doctors, insurers, pharmacies and patients > handle administrative tasks across systems > give practices the capacity of 3–5 additional full-time staff within 3 months, on average this is what AI solving real-world problems looks like.frozen9d ago1.4K37612024.04%
AI is everywhere now, but still hasn't made things we buy every day any cheaper Aonic Club uses AI to cut out the middlemen, so you pay exactly what the product costs a 32 g protein shake costs $1.61 at Aonic at Costco it's $2.30. that's 30% less membership is only $49 for the first year, then $149, and that's the only thing they make money onfrozen10d ago1.4K2416032.41%
why do I still have to call a restaurant just to book a table? with bland, you text it any business's website, and that business gets its own text line with an AI agent behind it it can answer questions, book appointments, and even take payment right in the chat and if you've already called, it remembers, so you don't start overfrozen8d ago1.4K1535011.75%
DigitalOcean just fixed the most annoying thing about running agents in prod agent fails a user, logs look fine, you later find out 400 users we affected > signals flags it in minutes > simulations tests the fix before it ships > a human approves, and the bug dies at user one all in previewfrozen11d ago1.3K22112012.67%
"I want my agent to buy things for me. I just don't want it to see my card details" Self just launched Agent Vault. your agent can pay and log in for you, but it never actually sees your card or password you can also set your own rules, like a spending cap or approving anything you set it up once, then use it across all your agentsfrozen8d ago1.3K2849023.26%
someone is about to vibecode an app and walk away with a DGX Spark Synteq is running a hackathon, > deploy an app on their platform, > QRT their post with proof it's live, and you're in. > everyone gets $20 in free credits to start if you've got a side project sitting at 90%, this is your signfrozen3d ago1.3K2916043.13%
why are you still paying a big LLM to do your agent's easiest jobs? this is a new kind of model from the Scale Down team. it reads your input once and points straight to the answer it is 200x faster, 467x cheaper than frontier LLMs, and it handles 1M tokens of text and imagesfrozen10d ago1.3K27112013.24%
building an ai product in 2026 means planning first, add browser automation, execution all before the actual product pine turned that stack into one sdk > create a cloud computer > give it a job in plain language > get the finished work back > it keeps going while you're away https://t.co/AptnVzdOHwfrozen14d ago1.1K1615022.09%
you already paid for ads, the product, & shipping then a chargeback takes the sale back Disputely is fighting those free all q4 you keep 100% of whatever they win22d ago1.1K1213011.49%
anyone building with AI agents knows the wait. every slow model call stacks up and the whole task drags most of that wait isn't even the chip thinking, it's the chip waiting on memory @general_compute is teaming up with @positron_ai to fix that: > GPUs read the prompt, > Positron's memory-first chips write the answerfrozen3d ago1.1K3129033.97%
I'd love a robot to clean my house, but every one I've seen looks too scary to leave alone with my stuff Voxel from Tuesday Labs is different. it's small, friendly, and a little too obsessed with tidying up > puts everything back where it belongs > grabs your book or a drink > finds the sunglasses you lost > waters your plants ships in 2028, and only 999 numbered Series-1 units will ever be made https://t.co/MnyrwVtipufrozen3d ago1.1K2938033.85%
who should own the AI your company runs on, you or one big provider? that's the big question at the Open Intelligence Summit today if one company controls your model, your data and everything your agent learns, you're stuck with them, even if a better or cheaper option comes along DigitalOcean's CEO thinks it should stay open, so you can switch any piece whenever you want. live at 1pm PTfrozen10d ago1.1K2014012.41%
if you scroll any brand's page right now, you can spot AI images in a second, they all look the same this is because image models don't know your brand, so they just guess Vyb starts with your brand instead. > enter your website, it pulls in your real logo, fonts and colors > uses your actual product photos, never redraws them > builds every piece on its own layer, so you can change just the headlinefrozen3d ago98526112024.16%
you finished building an app in a weekend and then you spend another week trying to get it online Synteq just launched Cloud Apps to take that part off your plate > deploy straight from GitHub, with one command, or by drag-and-drop > no servers to set up, no config files just got my $20 in free credit to try itfrozen3d ago9582137003.24%
made by human..is the new luxury13d ago779502000.90%
Vibe coders write apps. Engineers build systems.13d ago7681205002.21%
look closer at meta's coding numbers. > long-horizon coding: muse wonder wins > ultra-long engineering tasks: muse wonder wins, by a lot > writing a program from scratch: opus 5.5 still way ahead it's not that good at coding. it's better at handling larger tasks https://t.co/hvHJy2UQh211d ago5981111012.34%
it's becoming clear now.. > Microsoft is building its own AI chips instead of relying entirely on Nvidia > they now designed a Maia 300 chip to run AI workloads > plans to deploy it at scale in 2027 the ability to run AI cheaply will be one of the biggest advantages any AI company can have.4h ago497704002.21%
please don't use ai to generate memes they look ugly6d ago465614002.37%
cheap models are the worst thing to happen to AI startups. they used to charge $29/month for a wrapper. now users just use Haiku and do it themselves for free.14d ago370100000.27%
entering the weekend knowing monday i get to use the new Devin Desktop https://t.co/bYfMA5BYVW3h ago357301001.12%
paying for ChatGPT, Claude, Devin and Perplexity, and still googling the answer. https://t.co/eShFWYJwIm22d ago300603003.00%
tell them to pace the frontier and release 2 models in 6 days https://t.co/9UYeHDMELH25d ago251400001.59%
80 hours?? I'm not sure if they're hiring humans or agents https://t.co/ojDHaeNDgj14d ago241200000.83%
there is absolutely no reason to pay $200 in AI subscription to one provider. $100 into Claude and $100 OpenAI12d ago230201001.30%
after Opus 5.5 every other model feels like talking to customer support.15d ago227101000.88%
48% of people thought they were on a real video call @tavus griffin isn't a face stuck on a chatbot it watches, listens and talks at the same time > plays simon says > notices someone walk in the first model to pass the video turing test22d ago208200000.96%
Opus 5.5 turns Minecraft into a realistic Star Wars game, in the browser https://t.co/Se7YY2EEuk17d ago202200000.99%
if Sonnet 5.5 is almost better than Opus 5.5, what will Haiku 5.5 be like ?23d ago144000000.00%
Opus 5.5 has made everyone a better vibe coder than any other model before.4d ago105100000.95%
half my timeline says opus 5.5 got nerfed the other half says it's the best it's ever been.. who, what is the problem6d ago84100001.19%