[wbnapier]

I just asked an AI to select a HRC for my engine.

I gave it basic info, basically my engine.

Intended use: Street. 455/469, #48 Heads, '65 Tripower, 9.25:1, 4 speed, 3.55 Dif

Need vacuum for brakes.

It recommended:

COMP Cams Xtreme Energy XR282HR Hydraulic Roller Camshaft

Advertised duration: 282 degrees intake / 288 degrees exhaust
Duration at .050": 230 degrees intake / 236 degrees exhaust
Gross valve lift: 0.525" intake / 0.530" exhaust
Lobe separation angle: 110 degrees

I'm curious what you all think of that camshaft given the parameters I set out.

[wbnapier]

Follow up, I asked the AI for projected performance, it said:

Based on my experience, a well-tuned engine with this camshaft could produce approximately 450-500 horsepower at the flywheel and 500-550 lb-ft of torque, depending on factors such as the specific parts used and the engine's state of tune. This estimate is just a rough approximation, and your actual results may vary depending on several factors.

It's worth noting that the power and torque figures can be further improved by selecting the appropriate supporting components, such as the cylinder heads, intake manifold, carburetor, and exhaust system. Proper tuning and calibration of the engine management system is also essential to achieve the best possible performance.

[242177P]

Originally Posted by wbnapier

Follow up, I asked the AI for projected performance, it said:

Based on my experience ...

The AI doesn't have any experience. It takes things others have
posted, rewords it, and spits it right back at you. Did it go with
a tried and true recipe, or did it stumble upon a bunch of idiots
offering bad advice?

These AI programs are sophisticated search engines. 50 years
ago, it would've told you to run a 3/4 race cam.

[Drag Star Le Mans]

AI advice is like Fakebook advice.

[dataway]

Two things to consider.

Regardless of the source of the information (and 2421 is correct about the source) what is the accuracy of the information the AI provides. Searching the internet may indeed provide an accurate answer, also it may not. An experienced builder can tell you that ... oh snap ... you mean you could have just asked an experienced engine builder in the first place?

And .. you can be sure that the AI did not perform a single physics based calculation to come to the conclusion that it did. It accessed the results of other people/sources that did the actual calculations.

Very interesting experiment, would love to see it answer other questions ... how about the best way to repair an Endura bumper Or what is better, conventional or synthetic oil

[Murf]

Seems like that’s been a standard “go to” cam for a lot of vendors over the years. Ken recommended that cam or one that is very similar to me probably 30 yrs ago,
I have a combination very similar to yours. I have a Qjet instead of a tripower. Runs pretty good. If I was going to change it I’d probably go to a wider LSA.

I’m sure people who know more than I do will give more info than I can.

Murf

[1968GTO421]

Butler sells Comp cams but also

[1968GTO421]

Butler sells Comp cams and they offer a bunch of Comp grinds using the 112 lobe separation. Butler has been involved in racing Ponchos for years as well as selling parts. Members can get free advice from them using their experience.
Here are the Butler/Comp grinds:

https://butlerperformance.com/c-1272...om-grinds.html

(and most members already know Butler. I find the OP's posting interesting but am dubious of AI, Too many negative probabilities with it.)

(Botched post #7, kindly ignore. Thanks)

[Half-Inch Stud]

So, the AI answer is correct, and based on published information. The rub is that Unpublished information may be more correct.

[jhein]

According to the AI (actual intelligence) here on PY and from Butler, that would be an acceptable answer but maybe not necessarily the best answer. 469 is a lot of cubes and can take more cam than that. I have the Butler version of the 288HR on a 112 and it idles great with good vacuum. I wouldn't go any smaller and maybe even just a little bigger, especially for a manual with no AC (I don't know if your car has AC). But, I'm at 10:1 CR and maybe the bigger cam would want more than 9.25??

See, maybe this is one way that AI (artificial intelligence) will kill us all, by recommending unreasonably small cams. :

https://www.foxnews.com/tech/ai-expe...earth-will-die

[wbnapier]

I specified a streetable cam with lots of vacuum.

I asked the AI if it had access to Pontiac dyno data and it said yes, but only up to 2021, and only if it was published as text online. In theory, someone could create a standardized format for dyno sessions, listing the variables and numbers and let the AI crunch the data.

I asked the AI if he has access to David Vizard published works, and it replied yes, and that it factors in this knowledge.

[wbnapier]

Originally Posted by 1968GTO421

Butler sells Comp cams and they offer a bunch of Comp grinds using the 112 lobe separation. Butler has been involved in racing Ponchos for years as well as selling parts. Members can get free advice from them using their experience.
Here are the Butler/Comp grinds:

https://butlerperformance.com/c-1272...om-grinds.html

(and most members already know Butler. I find the OP's posting interesting but am dubious of AI, Too many negative probabilities with it.)

(Botched post #7, kindly ignore. Thanks)

I'm just having fun with the new AI tech and I really like Pontiac engines, so I wanted to play around.

I have no doubt that AI will only get better as time goes on.

[Tandyman]

WB , what company is behind this AI and how does one access it? It does sound like good entertainment .
Glenn

[wbnapier]

https://openai.com/

[jhein]

Originally Posted by wbnapier

I specified a streetable cam with lots of vacuum.

I asked the AI if it had access to Pontiac dyno data and it said yes, but only up to 2021, and only if it was published as text online. In theory, someone could create a standardized format for dyno sessions, listing the variables and numbers and let the AI crunch the data.

I asked the AI if he has access to David Vizard published works, and it replied yes, and that it factors in this knowledge.

I would assume it has access to posts on this forum too.

[Half-Inch Stud]

Originally Posted by jhein

I would assume it has access to posts on this forum too.

I assume the Clubhouse content is not accessed.

[jhein]

Originally Posted by Half-Inch Stud

I assume the Clubhouse content is not accessed.

Yeah, lots of talk about getting shafted over there, just not cam-shafted.

[Cliff R]

"I specified a streetable cam with lots of vacuum."

Depending on your goals for the project I would go wider on the LSA.

About 10 years ago I had a LONG discussion with Rodney Butler about camshafts. It's a LONG story and I'll save the readers here the details, but what came out of it and some emails back and forth was that Butler started offering WIDE LSA camshafts based on lobe profiles I'd been using here.

I'm not sitting here trying to take credit for any of that, could give a rats but about getting patted on the back at my age. However, whether or not they followed my advice or just accidentally dyno'd some wider LSA cams and started making a LOT more power without all that reversion/low vacuum at idle speed who knows.

The first pic below is from an engine very close to the one in discussion here. They used Butlers 236/242/114LSA camshaft and unported #16 heads. My only contrbution other than the cam recomendation was "tweak" his Q-jet for the dyno session.

The second pic is one of my favorites. The engine builder had the 455 on the dyno and it wasn't making nearly the power they expected plus pinging when they tried to put very much timing in it. Of course since I contributed to the Q-jet being used on it EVERYONE right down to the guy taking out the trash was blaming the carburetor. I told them to remove the XR276HR cam and give it a "hook shot" for the dumpster and put in the second cam shown on the dyno chart. Reluctantly they did and were nicely rewarded with smoother idle, imporoved throttle response and a BUTTLOAD more power.......AND......no more detonation!........FWIW.....

[Half-Inch Stud]

@CLIFF, 30 HP increase in the low rpm, yet amazing HP increase goingk north of 4400 rpm. Sure it wasn't a stinkbug covering the Secondary Jets huh.

I want to know all about that XR276 HR to know what induces crowbarred Low Timing ADV from Ping, and the low HP.
Yet surprising to see/know the OF delivering the goods with 114 LSA.

So, yea a mystery (for me) how the best cam parameters stuck on 106 LSA would compare.

High-Compr was a relative easy cam optimization compared to low Compr for low Octane power.

[Stan Weiss]

Originally Posted by Half-Inch Stud

@CLIFF, 30 HP increase in the low rpm, yet amazing HP increase goingk north of 4400 rpm. Sure it wasn't a stinkbug covering the Secondary Jets huh.

I want to know all about that XR276 HR to know what induces crowbarred Low Timing ADV from Ping, and the low HP.
Yet surprising to see/know the OF delivering the goods with 114 LSA.

So, yea a mystery (for me) how the best cam parameters stuck on 106 LSA would compare.

High-Compr was a relative easy cam optimization compared to low Compr for low Octane power.

Mark,
Yes the OF had a wider LSA than the 276. It also had more seat duration, more 0.050 duration and if I remember correctly more overlap.

Stan