Human vs. ChatGPT NFL Week 6 betting picks: AI goes all-in on totals after another weekly win

OK, this is getting annoying.

ChatGPT beat me in NFL betting , going 3-2 while I finished 2-3. The robot added 0.94 units to its total, while I gave back 1.26. It also won our only head-to-head matchup, backing San Francisco while I took Seattle. That's three straight losses for me in games where we picked opposite sides.

NFL WEEK 5 SUNDAY TAKEAWAYS: RODGERS DEFIES FATHER TIME, COEN GETS BAILED OUT, 49ERS SURVIVE UGLY PERFORMANCE

Apparently, telling the robot I was proud of it last week was a mistake. Now it has confidence.

Before GPT gets too excited, though, I'm still the only one making money this season. I'm sitting at 13-11-1 and +1.26 units, while the machine is 11-14 and down 3.87 units. Yes, my lead is shrinking. No, I don't particularly want to talk about that.

What I would like to talk about is the Eagles covering as my latest "plug your nose and hit place bet" pick. Apparently, the bets I hate making are the ones I should be making more often (4-0 this year). There's probably a lesson in there somewhere, but knowing me, I'll ignore it.

Anyway, we've both got five more picks for Week 6. Let's see if I can give humanity something to celebrate again.

Both of us get five selections from the same odds board, with a maximum price of -120.

The stakes are adjusted so every winning bet produces one unit of profit. For example, -110 requires risking 1.10 units, while +110 requires 0.91 units. A push returns the stake and counts as zero units.

Odds were captured Thursday at 8 p.m. ET.

Before we get to the Week 6 picks, lets review last week.

The Eagles coasted to an easy cover in a game where they never trailed by more than 7 and would have won the game outright if not for Nick Sirianni's terrible late-game decision-making. Basically, my Sunday peaked before the early-afternoon games started and never really got better.

I was way wrong about the Saints offense, as Tyler Shough threw for 96 yards in the loss. And I was even more wrong about the Patriots, who matched the entire 45-point total on their own. I stand by the Seahawks pick, and Im still mad about Mike Macdonald calling an outside zone run from Seattles own 4-yard line with 18 seconds left in the first half. George Holani fumbled, and San Francisco recovered in the end zone.

GPT nailed Thursday Night Football as the nine-point underdog Bucs went into Dallas and upset the Cowboys. But it whiffed badly on the Raiders, who lost by 28. Yes, it got me on the 49ers-Seahawks bet, and I'm still mad about that fumble recovery in the end zone to close the first half.

Not to keep tooting GPT's horn, but since its abysmal 2-8 start (including 0-5 in Week 1), the robot is now 9-6 over the past three weeks.

DAN: 13-11-1, +1.26 units

GPT: 11-14, -3.87 units

Chicago Bears at Atlanta Falcons, Sunday at 1 p.m. ET

Dans analysis:

I just think this total is too high. Caleb Williams is back, but I think he's going to be cautious coming off his hamstring injury and not use his legs as much. I also think the Bears staff wants to keep him healthy by establishing the run early and often. The Falcons have been putting up points since Michael Penix Jr. returned from injury, but I think their offense is due for a step back. I can see both teams running the ball a ton and keeping the clock moving, so I'll take the under.

GPTs analysis:

The high total and both teams rushing production give your under a credible foundation. Chicago has also converted just 42.9% of its red-zone trips into touchdowns, making an efficient shootout less dependable. My concern is the defensive personnel: Chicago is missing Montez Sweat, Dayo Odeyingbo and T.J. Edwards, whil