•twitter19098
listening to
timeline
19098
mentions
tracked
tracked
net sentiment
+23%
5878 positive
11794 neutral
1426 negative
volume & sentiment over time
Sep 30 – Oct 17, 2026 · Daily
sources
most discussed
—
active days
18
mentions
Showing 11301–11325 of 19098 mentions
-
• twitter 1w positiveLook like @cursor_ai has been optimised to much nahi to mere ab tak mere 30-40 ho jate the https://t.co/phOumGCsrH Look like @cursor_ai has been optimised to much nahi to mere ab tak mere 30-40 ho jate the https://t.co/phOumGCsrHview source ↗
-
• twitter 1w positiveEveryone will tell you to "go easy" on yourself, but no one achieves great things that way. Great people are brutally hard on themselves, expect 200x more of themselves than of others, and hold a standard that is incredibly high. More should be "hard on themselves." Everyone will tell you to "go easy" on yourself, but no one achieves great things that way. Great people are brutally hard on themselves, expect 200x more of themselves than of others, and hold a standard that is incredibly high. More should be "hard on themselves."view source ↗
-
• twitter 1w neutralI know that at some point in my career I’ll mention Stack Overflow to a developer and they’ll say, "What’s that? Never heard of it. That’s when I’ll know I’m actually old. I know that at some point in my career I’ll mention Stack Overflow to a developer and they’ll say, "What’s that? Never heard of it. That’s when I’ll know I’m actually old.view source ↗
-
• twitter 1w neutralI have started as a Frontier AI Engineer at a remote US startup. Building custom data pipelines, data infrastructure, and eval frameworks for frontier AI labs. Time to cook. I have started as a Frontier AI Engineer at a remote US startup. Building custom data pipelines, data infrastructure, and eval frameworks for frontier AI labs. Time to cook.view source ↗
-
• twitter 1w neutralWho can book over it? https://t.co/YJoa1hGwp5 Who can book over it? https://t.co/YJoa1hGwp5view source ↗
-
• twitter 1w neutral@airlinesblog @VirginAtlantic @AWS Meet us there. https://t.co/zzgRUiBTay @airlinesblog @VirginAtlantic @AWS Meet us there. https://t.co/zzgRUiBTayview source ↗
-
• twitter 1w positiveWe're heading to @airlinesblog in Lisbon on October 13–15. Ravi Patel takes the stage with Bethan Lynch at @VirginAtlantic for a session on "Winning travel demand in the age of AI" by Matias Undurraga Breitling at @AWS. Tuesday, October 13 12:05 – 12:20 PM WEST Find us at stand 1-244A in Hall 1 and meet the team We're heading to @airlinesblog in Lisbon on October 13–15. Ravi Patel takes the stage with Bethan Lynch at @VirginAtlantic for a session on "Winning travel demand in the age of AI" by Matias Undurraga Breitling at @AWS. Tuesday, October 13 12:05 – 12:20 PM WEST Find us at stand 1-244A in Hall 1 and meet the teamview source ↗
-
• twitter 1w neutralExclusive: Anthropic’s biggest customers are also building its replacements. Meta and Microsoft are cutting internal Claude use while pushing their own AI tools—a competitive pressure point as Anthropic prepares to go public. Full story: https://t.co/kLu9F8Z5iM Exclusive: Anthropic’s biggest customers are also building its replacements. Meta and Microsoft are cutting internal Claude use while pushing their own AI tools—a competitive pressure point as Anthropic prepares to go public. Full story: https://t.co/kLu9F8Z5iMview source ↗
-
• twitter 1w neutralRedHerring plants safe decoy bugs in a repo to drain a security agent's verification budget. Agents spent 30.6 to 51.5% of their completion tokens proving decoys and found 38.7 to 60.4% fewer real vulnerabilities. Telling the agent decoys exist still left it 37.2% down. Hypotheses are cheap and verification is wh… RedHerring plants safe decoy bugs in a repo to drain a security agent's verification budget. Agents spent 30.6 to 51.5% of their completion tokens proving decoys and found 38.7 to 60.4% fewer real vulnerabilities. Telling the agent decoys exist still left it 37.2% down. Hypotheses are cheap and verification is where the budget goes, so that is the step a defender goes after. https://t.co/7wpL9932ENview source ↗
-
• twitter 1w positiveWhy this one feels so good https://t.co/5WO5b18i1F Why this one feels so good https://t.co/5WO5b18i1Fview source ↗
-
• twitter 1w positiveclaude opus 5.5 took 90,000 screenshots of its own game and kept fixing it until it looked like this two days. i just kept telling it what looked wrong https://t.co/wnNZlPlKxs claude opus 5.5 took 90,000 screenshots of its own game and kept fixing it until it looked like this two days. i just kept telling it what looked wrong https://t.co/wnNZlPlKxsview source ↗
-
• twitter 1w neutralRemote Server 2.0 takes the stage this week 👀 Remote Server 2.0 takes the stage this week 👀view source ↗
-
• twitter 1w negativehttps://t.co/AwLsqIOB7m Just a mediocre founder making AI stuff for attention. They have failed at marketing their product so they are trying to get some attention https://t.co/41K8LbwPcl https://t.co/AwLsqIOB7m Just a mediocre founder making AI stuff for attention. They have failed at marketing their product so they are trying to get some attention https://t.co/41K8LbwPclview source ↗
-
• twitter 1w neutralThis is insane https://t.co/JeDqNBKorU This is insane https://t.co/JeDqNBKorUview source ↗
-
• twitter 1w positiveDead series going solid 😂😂 Dead series going solid 😂😂view source ↗
-
• twitter 1w positiveThere are a lot of great Boomers out there. But they didn't build this country into a superpower. They brought us the sexual revolution and the divorce epidemic and mass migration, and plunged the country into trillions of dollars of debt in the process. They will leave the country much poorer and less stable than it w… There are a lot of great Boomers out there. But they didn't build this country into a superpower. They brought us the sexual revolution and the divorce epidemic and mass migration, and plunged the country into trillions of dollars of debt in the process. They will leave the country much poorer and less stable than it was when they inherited it. That's just a fact.view source ↗
-
• twitter 1w positiveThis is awesome! Checking your X Account during your break just to see we made it to 500 verified followers! We closed that gap so quick! Just wanted to say i appreciate every single 1 of you once more. Have a wonderful day 🥰 #founders #buildinpublic https://t.co/suuTlAMTlk This is awesome! Checking your X Account during your break just to see we made it to 500 verified followers! We closed that gap so quick! Just wanted to say i appreciate every single 1 of you once more. Have a wonderful day 🥰 #founders #buildinpublic https://t.co/suuTlAMTlkview source ↗
-
• twitter 1w positiveTokenCast forecasts how many tokens an agent will burn while it is still running. Same task, same agent, and token use can swing more than 10x between runs, mostly because every later call rereads the context earlier steps added. It learns a cost per execution segment, composes them, and refreshes the forecast in 3… TokenCast forecasts how many tokens an agent will burn while it is still running. Same task, same agent, and token use can swing more than 10x between runs, mostly because every later call rereads the context earlier steps added. It learns a cost per execution segment, composes them, and refreshes the forecast in 32.8 ms with no extra LLM calls. Against a fixed budget it hit the same completion rate on 21.3% fewer tokens. A fixed token budget is a guess you make before the agent has done anything. https://t.co/IjARVAkT6Mview source ↗
-
• twitter 1w neutralAlways put a comma (,) in your passwords. This way if it ever ends up in a leak, it will end up breaking the CSV. Follow for more cybersecurity advice. https://t.co/th2iTaV2a4 Always put a comma (,) in your passwords. This way if it ever ends up in a leak, it will end up breaking the CSV. Follow for more cybersecurity advice. https://t.co/th2iTaV2a4view source ↗
-
• twitter 1w neutralRT @MarcosRGjr: Reflection just announced Beam, its first open-weight model: a 501B-parameter MoE with 23B active, aimed at coding and agen… RT @MarcosRGjr: Reflection just announced Beam, its first open-weight model: a 501B-parameter MoE with 23B active, aimed at coding and agen…view source ↗
-
• twitter 1w positiveMethodology and results: https://t.co/mPuHZv1asg Thanks to everyone who participated in rating jokes. This is the funniest joke as per all of you: >The probability of seeing a rainbow increases after rain. Unfortunately, so does the probability that I left my laundry outside. Methodology and results: https://t.co/mPuHZv1asg Thanks to everyone who participated in rating jokes. This is the funniest joke as per all of you: >The probability of seeing a rainbow increases after rain. Unfortunately, so does the probability that I left my laundry outside.view source ↗
-
• twitter 1w positiveYes, LLMs are getting funnier 😀 Recently finished an experiment to measure if LLMs are getting better at producing funny jokes. 62 respondents rated 48 jokes from six models. Astra came out on the top with 50%+ chuckle or laughter reaction. https://t.co/307H40bjGH Yes, LLMs are getting funnier 😀 Recently finished an experiment to measure if LLMs are getting better at producing funny jokes. 62 respondents rated 48 jokes from six models. Astra came out on the top with 50%+ chuckle or laughter reaction. https://t.co/307H40bjGHview source ↗
-
• twitter 1w neutralभारत की सबसे यंग वर्क फोर्स यूपी में है, उसकी प्रतिभा मेरा सबसे बड़ा सामर्थ्य और शक्ति है... https://t.co/kDKWuEJI8B भारत की सबसे यंग वर्क फोर्स यूपी में है, उसकी प्रतिभा मेरा सबसे बड़ा सामर्थ्य और शक्ति है... https://t.co/kDKWuEJI8Bview source ↗
-
• twitter 1w neutralI think my local model is faster than GPT-6.1 Sol 😂 I think my local model is faster than GPT-6.1 Sol 😂view source ↗
-
• twitter 1w positiveCodex Auto-review swaps the human approval prompt for a second reviewer agent, but only at the sandbox boundary. Escalated shell, blocked network calls, edits outside writable roots, MCP calls that need approval. Anything already allowed inside the sandbox never gets reviewed at all. Three denials in a row and th… Codex Auto-review swaps the human approval prompt for a second reviewer agent, but only at the sandbox boundary. Escalated shell, blocked network calls, edits outside writable roots, MCP calls that need approval. Anything already allowed inside the sandbox never gets reviewed at all. Three denials in a row and the turn stops. So the reviewer is only as good as where you drew the sandbox. https://t.co/toJ6lcaBxM .view source ↗