<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=Windows-1252">
<style type="text/css" style="display:none;"> P {margin-top:0;margin-bottom:0;} </style>
</head>
<body dir="ltr">
<div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);" class="elementToProof">
I'd like to pick up on one point Gary mentions. There are plenty of applications of AI that don't require replicating the human mind.</div>
<div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);" class="elementToProof">
<br>
</div>
<div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);" class="elementToProof">
However, replicating the human mind is a worthy goal in and of itself, independent of applications. If we accept certain positions in the philosophy of mind, such as functionalism and materialism, then replicating the human mind in a machine is a scientific
quest to understand ourselves in the most precise way possible: a computable model. This is the same motivation shared by psychology and other social sciences.<br>
</div>
<div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);" class="elementToProof">
<br>
</div>
<div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);" class="elementToProof">
A complete computable model of the human mind should also explain all relevant observations from neuroscience, even if it is not literally a whole-brain model.<br>
</div>
<div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);" class="elementToProof">
<br>
</div>
<div id="appendonsend"></div>
<div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<hr tabindex="-1" style="display:inline-block; width:98%">
<div id="divRplyFwdMsg" dir="ltr"><font style="font-size: 11pt; color: rgb(0, 0, 0);" face="Calibri, sans-serif"><b>From:</b> Connectionists <connectionists-bounces@mailman.srv.cs.cmu.edu> on behalf of Gary Marcus <gary.marcus@nyu.edu><br>
<b>Sent:</b> Thursday, June 8, 2023 7:35 AM<br>
<b>To:</b> Grossberg, Stephen <steve@bu.edu><br>
<b>Cc:</b> connectionists@cs.cmu.edu <connectionists@cs.cmu.edu><br>
<b>Subject:</b> [EXTERNAL] Re: Connectionists: Sentient AI Survey Results</font>
<div> </div>
</div>
<div class="BodyFragment"><font size="2"><span style="font-size:11pt">
<div class="PlainText">I would add to the below that there are multiple reasons why current models are neither trustworthy nor reliable (e.g., also because they lack world models and adequate robust processes for reasoning) but Stephen has really hit the nail
on the head here. AI doesn’t need to *replicate* the human mind, but current systems very much do “omit key processes that make humans intelligent”, and the fact that they are untrustworthy and unreliable follows directly from there.<br>
<br>
We must consider alternative approaches if we are to rectify the untrustworthiness and unreliability.<br>
<br>
> On Jun 8, 2023, at 00:27, Grossberg, Stephen <steve@bu.edu> wrote:<br>
> <br>
> Deep Learning, and related models, are biologically impossible and omit key processes that make humans intelligent, in addition to being UNTRUSTWORTHY (because they are NOT EXPLAINABLE) and UNRELIABLE (because they experience CATASTROPHIC FORGETTING).<br>
<br>
</div>
</span></font></div>
</body>
</html>