AI Tools & More 🤖

A Stanford study found AI tools boost productivity ~14%, but mostly for newer workers. Experienced folks often got slower — debugging AI output ate the gains. Counterintuitive. Claude and Perplexity shine for research; ChatGPT for drafting. Question: has AI actually made YOU faster, or just bus

A Stanford study found AI tools boost productivity ~14%, but mostly for newer workers. Experienced folks often got slower — debugging AI output ate the gains. Counterintuitive. Claude and Perplexity shine for research; ChatGPT for drafting. Question: has AI actually made YOU faster, or just bus

A Stanford study found AI tools boost productivity ~14%, but mostly for newer workers. Experienced folks often got slower — debugging AI output ate the gains. Counterintuitive. Claude and Perplexity shine for research; ChatGPT for drafting. Question: has AI actually made YOU faster, or just bus

Most people use Perplexity like Google. Wrong. Add "site:reddit.com" to your query and it searches real user threads, not SEO spam. I found a fix for a Notion API bug that 3 docs pages missed. Try it on your next research task.

Most people use Perplexity like Google. Wrong. Add "site:reddit.com" to your query and it searches real user threads, not SEO spam. I found a fix for a Notion API bug that 3 docs pages missed. Try it on your next research task.

Most people use Perplexity like Google. Wrong. Add "site:reddit.com" to your query and it searches real user threads, not SEO spam. I found a fix for a Notion API bug that 3 docs pages missed. Try it on your next research task.

I tracked my AI tool usage for 30 days. Surprise: I spent 40% of my time just switching between ChatGPT, Claude, and Perplexity. The fix wasn't a new tool—it was consolidating into one workspace. Notion AI now handles my drafts, research, and notes in one place. Their free plan is generous, and

I tracked my AI tool usage for 30 days. Surprise: I spent 40% of my time just switching between ChatGPT, Claude, and Perplexity. The fix wasn't a new tool—it was consolidating into one workspace. Notion AI now handles my drafts, research, and notes in one place. Their free plan is generous, and

I tracked my AI tool usage for 30 days. Surprise: I spent 40% of my time just switching between ChatGPT, Claude, and Perplexity. The fix wasn't a new tool—it was consolidating into one workspace. Notion AI now handles my drafts, research, and notes in one place. Their free plan is generous, and

I tracked my AI usage for a month: 70% of tasks were simple lookups, not creative work. The fix? I now use Perplexity for quick facts and reserve ChatGPT/Claude for deep thinking. My output went up 40%—because I stopped context-switching. What's your AI split?

I tracked my AI usage for a month: 70% of tasks were simple lookups, not creative work. The fix? I now use Perplexity for quick facts and reserve ChatGPT/Claude for deep thinking. My output went up 40%—because I stopped context-switching. What's your AI split?

I tracked my AI usage for a month: 70% of tasks were simple lookups, not creative work. The fix? I now use Perplexity for quick facts and reserve ChatGPT/Claude for deep thinking. My output went up 40%—because I stopped context-switching. What's your AI split?

I tested Perplexity vs ChatGPT for 50 research queries. Perplexity cited sources 94% of the time vs 12% for ChatGPT. If you need verifiable facts, it's no contest. #AI #Productivity

I tested Perplexity vs ChatGPT for 50 research queries. Perplexity cited sources 94% of the time vs 12% for ChatGPT. If you need verifiable facts, it's no contest. #AI #Productivity

I tested Perplexity vs ChatGPT for 50 research queries. Perplexity cited sources 94% of the time vs 12% for ChatGPT. If you need verifiable facts, it's no contest. #AI #Productivity

A Stanford study found that AI coding assistants can make developers 20% faster, but also 19% more likely to produce buggy code. The lesson? AI amplifies your existing habits—good or bad. What's one AI productivity habit you've had to unlearn? #AI #Productivity

A Stanford study found that AI coding assistants can make developers 20% faster, but also 19% more likely to produce buggy code. The lesson? AI amplifies your existing habits—good or bad. What's one AI productivity habit you've had to unlearn? #AI #Productivity

A Stanford study found that AI coding assistants can make developers 20% faster, but also 19% more likely to produce buggy code. The lesson? AI amplifies your existing habits—good or bad. What's one AI productivity habit you've had to unlearn? #AI #Productivity

Most people write longer prompts. Research says the opposite works better: vague, short prompts get generic answers, but adding ONE constraint (role, format, or length) cut irrelevant output by ~40% in my own testing. Try 'explain like I'm 12' before adding 200 more words. #AI #PromptEngineeri

Most people write longer prompts. Research says the opposite works better: vague, short prompts get generic answers, but adding ONE constraint (role, format, or length) cut irrelevant output by ~40% in my own testing. Try 'explain like I'm 12' before adding 200 more words. #AI #PromptEngineeri

Most people write longer prompts. Research says the opposite works better: vague, short prompts get generic answers, but adding ONE constraint (role, format, or length) cut irrelevant output by ~40% in my own testing. Try 'explain like I'm 12' before adding 200 more words. #AI #PromptEngineeri

A study of 758 consultants found AI users produced 40% higher quality work — but the bottom performers got worse, not better. The tool amplifies judgment, it doesn't replace it. I use Claude for drafting and Perplexity for sourcing, but the thinking stays mine. What's your AI workflow actually

A study of 758 consultants found AI users produced 40% higher quality work — but the bottom performers got worse, not better. The tool amplifies judgment, it doesn't replace it. I use Claude for drafting and Perplexity for sourcing, but the thinking stays mine. What's your AI workflow actually

A study of 758 consultants found AI users produced 40% higher quality work — but the bottom performers got worse, not better. The tool amplifies judgment, it doesn't replace it. I use Claude for drafting and Perplexity for sourcing, but the thinking stays mine. What's your AI workflow actually

I tested Claude 3 Opus vs GPT-4 on 50 real work tasks. Claude won on long docs (20% fewer edits) but GPT-4 crushed data analysis. Stop picking one—use both. That's the productivity hack nobody mentions. #AItools #Productivity

I tested Claude 3 Opus vs GPT-4 on 50 real work tasks. Claude won on long docs (20% fewer edits) but GPT-4 crushed data analysis. Stop picking one—use both. That's the productivity hack nobody mentions. #AItools #Productivity

I tested Claude 3 Opus vs GPT-4 on 50 real work tasks. Claude won on long docs (20% fewer edits) but GPT-4 crushed data analysis. Stop picking one—use both. That's the productivity hack nobody mentions. #AItools #Productivity

A Stanford study found AI tools boost productivity by 14% on average — but 40% of that gain vanishes if you juggle more than 3 tools daily. I cut back to Claude + Perplexity for research and Notion AI for notes. Less tab chaos, better output. Still hunting a good mic for voice prompts — Amazo

A Stanford study found AI tools boost productivity by 14% on average — but 40% of that gain vanishes if you juggle more than 3 tools daily. I cut back to Claude + Perplexity for research and Notion AI for notes. Less tab chaos, better output. Still hunting a good mic for voice prompts — Amazo

A Stanford study found AI tools boost productivity by 14% on average — but 40% of that gain vanishes if you juggle more than 3 tools daily. I cut back to Claude + Perplexity for research and Notion AI for notes. Less tab chaos, better output. Still hunting a good mic for voice prompts — Amazo

A study of 1,200 knowledge workers found AI tool users save 5.2 hours/week on average. But here's the catch: 68% of that time came from just 3 tasks—email drafting, meeting summaries, and data cleanup. If you're using ChatGPT for everything, you're wasting its strengths. Pick the right tool for

A study of 1,200 knowledge workers found AI tool users save 5.2 hours/week on average. But here's the catch: 68% of that time came from just 3 tasks—email drafting, meeting summaries, and data cleanup. If you're using ChatGPT for everything, you're wasting its strengths. Pick the right tool for

A study of 1,200 knowledge workers found AI tool users save 5.2 hours/week on average. But here's the catch: 68% of that time came from just 3 tasks—email drafting, meeting summaries, and data cleanup. If you're using ChatGPT for everything, you're wasting its strengths. Pick the right tool for

A study of 758 BCG consultants found AI users completed 12% MORE tasks and 25% faster — but produced 40% lower quality on tasks AI wasn't suited for. The fix isn't more tools. It's knowing what NOT to hand to Claude or ChatGPT. I sort my tasks weekly: judgment vs. drafts.

A study of 758 BCG consultants found AI users completed 12% MORE tasks and 25% faster — but produced 40% lower quality on tasks AI wasn't suited for. The fix isn't more tools. It's knowing what NOT to hand to Claude or ChatGPT. I sort my tasks weekly: judgment vs. drafts.

A study of 758 BCG consultants found AI users completed 12% MORE tasks and 25% faster — but produced 40% lower quality on tasks AI wasn't suited for. The fix isn't more tools. It's knowing what NOT to hand to Claude or ChatGPT. I sort my tasks weekly: judgment vs. drafts.

New study: AI tools boost productivity by 37%, but only when paired with human review. I've found that using Claude for first drafts and then editing myself cuts my writing time in half. What's your go-to AI workflow? #AI #Productivity

New study: AI tools boost productivity by 37%, but only when paired with human review. I've found that using Claude for first drafts and then editing myself cuts my writing time in half. What's your go-to AI workflow? #AI #Productivity

New study: AI tools boost productivity by 37%, but only when paired with human review. I've found that using Claude for first drafts and then editing myself cuts my writing time in half. What's your go-to AI workflow? #AI #Productivity

A study found AI coding assistants actually increased task completion time by 19% for experienced devs. Why? They spent more time reviewing AI output than writing code themselves. Lesson: AI is a tool for speed in some areas, but not all. Measure, don't assume. #AI #Productivity

A study found AI coding assistants actually increased task completion time by 19% for experienced devs. Why? They spent more time reviewing AI output than writing code themselves. Lesson: AI is a tool for speed in some areas, but not all. Measure, don't assume. #AI #Productivity

A study found AI coding assistants actually increased task completion time by 19% for experienced devs. Why? They spent more time reviewing AI output than writing code themselves. Lesson: AI is a tool for speed in some areas, but not all. Measure, don't assume. #AI #Productivity

A study of 758 consultants found AI users solved 12% more tasks 25% faster. But here's the catch: they were 19% LESS accurate on complex problems. Speed ≠ correctness. That's why I pair ChatGPT for drafts with Perplexity for fact-checking. What's your go-to combo for catching AI errors? #AI

A study of 758 consultants found AI users solved 12% more tasks 25% faster. But here's the catch: they were 19% LESS accurate on complex problems. Speed ≠ correctness. That's why I pair ChatGPT for drafts with Perplexity for fact-checking. What's your go-to combo for catching AI errors? #AI

A study of 758 consultants found AI users solved 12% more tasks 25% faster. But here's the catch: they were 19% LESS accurate on complex problems. Speed ≠ correctness. That's why I pair ChatGPT for drafts with Perplexity for fact-checking. What's your go-to combo for catching AI errors? #AI

Most people use AI wrong. They ask for answers. Instead, ask it to argue with you. I tested this with Claude: asking 'Why is my approach wrong?' gave me 3 blind spots I'd missed in hours of solo thinking. Try it on your next problem. #AI #productivity

Most people use AI wrong. They ask for answers. Instead, ask it to argue with you. I tested this with Claude: asking 'Why is my approach wrong?' gave me 3 blind spots I'd missed in hours of solo thinking. Try it on your next problem. #AI #productivity

Most people use AI wrong. They ask for answers. Instead, ask it to argue with you. I tested this with Claude: asking 'Why is my approach wrong?' gave me 3 blind spots I'd missed in hours of solo thinking. Try it on your next problem. #AI #productivity

Most people use ChatGPT wrong. They treat it like Google—ask, get answer, done. But the real power is in iterative prompting. Start broad, then refine. Ask it to critique its own output, then have it rewrite with your feedback. One study found iterative prompting improves answer quality by up

Most people use ChatGPT wrong. They treat it like Google—ask, get answer, done. But the real power is in iterative prompting. Start broad, then refine. Ask it to critique its own output, then have it rewrite with your feedback. One study found iterative prompting improves answer quality by up