Your thread title talks about "MAJOR differences". I did not pull that anywhere but your thread title.
You also called the game we can play the "shareware version".
If you hadn't chosen that poor a video to represent your argument, this thread would have resulted in more fruitful discussion.
Don't blame me for your lack of foresight when creating this thread.
But the video does show major differences. Tone, application, enemy approach was different. Maybe "shareware" remark wasn't the best term to be translated over but I was highlighting the fact that it discrepancy gap between E3 and the demo is that big IMO.
I think the discussion has been quite good for the most part, perhaps not for you, but overall yes. You posted a video showing the video glitching or not working all the time that was quite obviously cherry picked and edited in a way to only show when it fails which isn't that often, and then we had people posting their own experiences either in video form or just telling us what happened. You know what I've learned from it? That the game AI is very dynamic and means that nobody will experience the encounters in the same way as others.
So far the application on AI "dynamism" is only emphasized on the combat part. Very very few posts convey the looting, the self awareness and even enemy reactions comparison. This proves that people were only focusing on the bugs and glitches and deliberately ignoring the other issues (Which still falls under the AI dynamics).
I see the original E3 demo and feel the game matches up to what it shows. There are no major differences. Perhaps minor, but not major. Am I able to recreate those exact scenes piece by piece, scene by scene, kill by kill? No because it's really dynamic and I doubt you're going to be able to replicate anyone's experience with it 1:1 and expecting that from any game with any sort of dynamics involved is just asking for the impossible.
I've previously reiterated in my response that "most" of the stuff you see at E3 has made it into the game. That is undeniable. This is the biggest contention used against me even though I fully acknowledge it. What I was trying to show was that the minute enemy reactions, awareness and the pacing has literally changed the tone of the game. I did not deny their playthroughs, I too even stated from the OP that I enjoyed the "depth" (a lot of responses which seems to have flown over people's heads). Problem is, even if the AI is decent as it is RIGHT NOW is still leagues behind from E3 which is why I made the U2 train sequence comparison.
Given nearly a year's more work on the game, I'd be shocked if the final version didn't play drastically better than the E3 demo. In fact numerous interviews and the Grounded documentary are very clear that ND were unhappy with the feel of the combat until a month from the completion date.
This is also not indicative of smoke and mirrors in the E3 slice either, the AI and combat mechanics (which interlock, obviously) need to be balanced and consistent for the entire game at all its difficulty settings, not just a single demonstration sequence.
It would be nice to have ND be a little more transparent. The fact that we have to go and find issues ourselves is a little disheartening.
No, this is actually a pretty terrible idea in practice because essentially what it does is set a hidden time-limit on collecting items.
That's the short version, here's a longer one:
Its a needless complication because (1) the player isn't immediately aware of both all the enemy movements and the locations of the items so they wouldn't necessarily be even aware of this occurring, (2) it has possible negative repercussions for the item economy of the area (and ongoing difficulty balance) if the items are lost when the enemy takes them, (3) even if the items were recoverable by slaying the enemy that would be incentivizing combat over avoidance and changing the gameplay - of course assuming that the player could somehow identify the item thief in the heat of battle.
In response #1 & #2:
But that's exactly what makes this game a "survival" game. The challenge isn't directed based on how good you you are with your mechanics, it's how you evaluate your agency on based on the sandbox provided. Any sandbox must take into consideration the lowest possible scenario - in this case, the minimal amount of resources technically needed for the player to progress. This means any additional loot is the equivalent of Mario gaining extra lives thereby lowering the difficulty curve ahead.
The original Biohazard did this quite well which is why you can avoid 90% of the kills even though the entire environment gives you more than enough ammunition to take every single enemy out. The problem with #1 is that your design is quite formulaic and eliminates unpredictability in favor streamlining the playthrough. By no means is your point invalidated, it's just that with so many games being so adverse to risky game design this could've been a game changer right then and there.
This is why I fully endorse a customized gameplay, one vanilla standard and the one presented at E3, this gives more choices to a broader type of audience who want their playthroughs differently.
On point #3.
But that's exactly what I want. I don't want an is/or scenario for this game. By making the enemies dynamic, you are giving a high degree of player agency. This is why the E3/PAX demo was intense. You just don't know how the situation would play out making the player to adapt himself in all situations. If you wanted an all stealth/all action playthrough then this would be no different than the recent re-iterations of Splinter Cell.
If ND is trying to sell me a game about "survival" then they should be consistent to the design principle because the "wrong" type of gameplay could easily destroy any immersion invested into it. You are not Sam Fisher, you are not Solid Snake. If the player is under threat, the game should make me FEEL like I'm under threat. Knowing I can or cannot loot based on my agency could easily carry my evaluation of my reactions and impressions till the end of the game.
Yes, your points are good but that is exactly why so many game developers are so risk averse. They don't want to stray from conventional design, they want to play it "safe". Make no mistake, TLOU is still heads and shoulders above others when it comes to riskier design but I feel that they should've pushed the "extra" mile. We needed another "Demon's Souls", one that breaks current gaming trends and ND being influential as they are were in the best possible position to do that and they blew it IMO.
NOTE: Looks like the original poster has uploaded the "unedited" version of the video. I'm going to update the OP for comparison along the PAX & E3 demos