Quinn Batten

I asked three different data leaders the same question over the past week: how fresh does your data actually need to be? I got three completely different answers. The first, at a healthtech company, needs sub-minute. Their operational database and their analytics layer feed the same workflows, and any gap between the two shows up as an inconsistency a clinician might see. The second runs monitoring on physical equipment. 5 minutes is the absolute ceiling, and ideally it is under one. The third runs internal analytics for a lean team. He told me 15 to 20 minutes is perfectly fine, and that probably would work for some of their agentic use cases. Most latency requirements are never traced. Someone writes "real-time" into an evaluation doc because it sounds rigorous, the vendor prices against it, and nobody asks what breaks at minute 6. The honest question sits downstream: what does the data feed, and what does it cost when it goes stale. For years, "real-time" could mean fifteen minutes and nobody got hurt. The slack was big enough that the imprecision never cost you - one dashboard, one report, one batch job, fifteen minutes covered all of it. You could write "real-time", mean almost anything, and be right. Agents close that gap. When an agent is acting on your data instead of a human, 70 seconds and 6 minutes are two different products. One catches the stale record before it acts. The other acts on it, and now you're cleaning up a decision instead of refreshing a chart. That's the shift we're seeing across the board So the rigor that never mattered suddenly does. Not "is it real-time" but what specifically breaks at 70 seconds, and what breaks at minute 6. That answer is a spec. "Real-time" is just a feeling that sounded rigorous in a doc. If you own a data platform: has anyone actually traced your strictest latency requirement lately - or is it still a word someone wrote a long time ago?

"ChatGPT is now Codex" ok sure I guess. But where did all the chats go that I started on my phone? And from my old ChatGPT app? Answer: Go to the left sidebar in the new desktop app and click "Chat." That brings up a little popup. Click "See all" on the bottom of that.

"ChatGPT is now Codex" ok sure I guess. But where did all the chats go that I started on my phone? And from my old ChatGPT app? Answer: Go to the left sidebar in the new desktop app and click "Chat." That brings up a little popup. Click "See all" on the bottom of that.

"ChatGPT is now Codex" ok sure I guess. But where did all the chats go that I started on my phone? And from my old ChatGPT app? Answer: Go to the left sidebar in the new desktop app and click "Chat." That brings up a little popup. Click "See all" on the bottom of that.

Chamath reveals his company's AI token costs are doubling every 45 days but productivity is only up 5% "I sat down with my CTO today, I said how are we doing on token spend. And he said the most incredible thing, he said right now, our token costs are doubling every 45 days. I said well what is the downstream productivity? And he said maybe 5% max." "So my costs are doubling every 45 days, my upside is essentially flat. He said honestly, what we're finding out is that you need to use a lot more tokens to get to this next iteration of improvement because we've effectively already asymptoted." "We're going to take a step back and try to figure out what to do. I don't know how many other companies will actually go through this reckoning now, but the point is everybody in the next three or four years will for sure go through it." "I suspect that if you can get out now, you should get out now before all of that starts to seep into the water table. Because I think that's probably what allows you to get out at a huge price and raise a huge amount of money."