Repository navigation
Expand file tree
/
Copy pathindex.html
More file actions
265 lines (261 loc) · 16.2 KB
/
Copy pathindex.html
File metadata and controls
265 lines (261 loc) · 16.2 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
<!DOCTYPE html>
<html lang="en-GB">
<head>
<title>Floating Octothorpe</title>
<meta charset="utf-8" />
<meta name="Author" content="Floating Octothorpe" />
<meta name="viewport" content="width=device-width, initial-scale=1">
<link rel="stylesheet" href="/css/main.css">
<link rel="icon" type="image/png" href="/img/favicon.png" />
</head>
<body>
<header>
<a href="/"><img class="logo" src="/img/logo.png" alt="Floating Octothorpe logo"/></a>
<h1 class="site-title">Floating Octothorpe</h1>
<nav>
<ul>
<li><a href="/">Latest</a></li>
<li><a href="/archive.html">Archive</a></li>
<li><a href="/about.html">About</a></li>
</ul>
</nav>
</header>
<article>
<h1>Thoughts on AI</h1>
<time datetime="2026-05-30">30 May 2026</time>
<p>It's been a minute since I last added anything to this site... well six years
to be precise. Since then technology, and more specifically AI, has changed
significantly. Large Language Models (LLMs) have gone from an academic topic to
being the zeitgeist of recent years.</p>
<p>Unlike a lot of posts on this site, this post is going to speculate on the
future; specifically where I think AI is going in the next few years. However,
like other posts on this site, this post is entirely my personal opinion; and,
like most future predictions, will likely look foolish in hindsight. With that
said, here is why I think the adoption of AI is going to be impactful; as well
as slower, messier and more constrained by economics, accountability, and
institutions than many people currently expect.</p>
<h2>Selling Shovels</h2>
<blockquote>
<p>"During a gold rush, sell shovels"</p>
</blockquote>
<p>Back in 2017, google introduced <a href="https://en.wikipedia.org/wiki/Transformer_(deep_learning)">transformer architecture</a> to
artificial neural networks with their paper <a href="https://papers.nips.cc/paper_files/paper/2017/file/3f5ee243547dee91fbd053c1c4a845aa-Paper.pdf">Attention Is All You
Need</a>. This opened the door for heavy
parallelisation during model training, something GPUs are well suited to. Since
then, Nvidia has rapidly grown as a company, overtaking Apple and Microsoft to
become the biggest company globally by market capitalisation. Personally I
can't see this changing any time soon, while many AI business models are
questionable at best, Nvidia are currently upstream of most other companies and
selling to a captive market. Longer term I think there are three main things
that could change this dynamic:</p>
<ol>
<li>The demand for AI workloads falls dramatically</li>
<li>New approaches to training supplant or diminish the need for GPUs</li>
<li>Competing companies (likely AMD) catch up or surpass Nvidia</li>
</ol>
<p>Until then Nvidia will continue to demonstrate that selling infrastructure is a
more reliable way to make profit than developing AI itself. Longer term I think
the competition will catch up, but chip development is a slow process.</p>
<h3>Predictions</h3>
<ul>
<li><strong>1 Year</strong>: Nvidia remain the largest global company by market cap.</li>
<li><strong>2 Years</strong>: Nvidia increase their lead over rival technology companies.</li>
<li><strong>5 Years</strong>: Competitors begin to offer viable alternatives to Nvidia, who
while still a dominant player in the GPU market, no longer hold
the crown as the world's largest company.</li>
</ul>
<h2>AI and Accountability</h2>
<p>I came across an old quote from an IBM training manual a couple of years ago:</p>
<blockquote>
<p>"A computer can never be held accountable, therefore a computer must never
make a management decision."</p>
</blockquote>
<p>The original quote is from 1979, and to me is strikingly relevant today.</p>
<p>The capability of LLMs, diffusion models, and generative artificial
intelligence has undeniably made huge advances in recent years. Do I think this
is going to profoundly change many professions? Absolutely. However, do I think
this is going to lead to mass redundancies? No, at least not for the
foreseeable future. Ultimately you can't hold an LLM accountable. <a href="https://en.wikipedia.org/wiki/Mata_v._Avianca,_Inc.">Mata v.
Avianca Inc.</a> is an early example of this, at some point in the
chain, a human is still accountable for the decision being made.</p>
<p>Self driving cars are another interesting case in point. The <a href="https://en.wikipedia.org/wiki/DARPA_Grand_Challenge">DARPA Grand
Challenge</a> started in 2004 as a competition to encourage
research and development in autonomous vehicles. This did just that, and within
a few years there was increased optimism that autonomous vehicles would become
commonplace. CGP Grey's video <a href="https://www.youtube.com/watch?v=iHzzSao6ypE">Traffic Has a Perfect Solution, Humans Won't Use
It</a>, is an interesting snapshot in time reflecting this. While
fundamentally the video does highlight real problems and solutions provided by
autonomous vehicles, it's too easy to come away with the impression that
everything is going to change in the next <em>few</em> years; I certainly felt that
way when I originally watched the video. With the benefit of hindsight progress
has certainly been slower than I would have thought 10 years ago. This has not
been helped by companies like Uber who were sharply criticized after the <a href="https://en.wikipedia.org/wiki/Death_of_Elaine_Herzberg">first
fatal collision</a> involving a self-driving car. With that said,
progress has been made and Waymo now operate a robotaxi service in a handful of
U.S. states, and plan to continue expanding their service.</p>
<p>Fundamentally, AI being used as a form of automation is inevitable when there
are clear economic incentives, however progress is often slower than people
first anticipate; and issues around accountability are at the heart of many of
the hurdles facing adoption.</p>
<h3>Predictions</h3>
<ul>
<li><strong>1 Year</strong>: LLMs like ChatGPT and Claude will increasingly be adopted in
workplace settings, with just under half of knowledge workers
regularly using some form of LLM. </li>
<li><strong>2 Years</strong>: Fully replacing roles using LLMs will have been attempted by
multiple companies, however attempts are almost universally
unsuccessful, with many leading to significant reputational
damage.</li>
<li><strong>5 Years</strong>: Many office-based roles use some form of AI on a regular basis,
however it is still very much a tool to be used instead of a
replacement for a worker. Unemployment rates in countries like
the U.S. and U.K. will not be significantly affected by AI,
remain at or below 5%.</li>
</ul>
<h2>Consciousness and Containment</h2>
<p>While I think there are legitimate concerns around how AI will impact job
markets, I'm much less sold on it becoming an existential threat to humanity.
On 30 May 2023 a <a href="https://aistatement.com/">Statement on AI Risk</a> was published:</p>
<blockquote>
<p>Mitigating the risk of extinction from AI should be a global priority
alongside other societal-scale risks such as pandemics and nuclear war.</p>
</blockquote>
<p>Among the list of signatories are many notable figures including Sam Altman,
Bill Gates, and Dario Amodei. Personally I feel the statement is too alarmist
and overplays the capability of a hypothetical future AI. Typically the logic
behind such statements is we will eventually produce an artificial general
intelligence (AGI), which can subsequently accelerate AI development and
produce a superintelligence that may or may not be aligned to humanity. At the
extreme you end up with beliefs like <a href="https://en.wikipedia.org/wiki/Singularitarianism">Singularitarianism</a>.
While definitely interesting ideas to muse over, they very quickly step over
the line of reality and into science fiction.</p>
<p>Another plot point that is increasingly crossing over from science fiction is
artificial consciousness. Chatbots in particular are easy to anthropomorphise,
something that dates back over 60 years to early chatbots like <a href="https://en.wikipedia.org/wiki/ELIZA">ELIZA</a>.
Modern LLMs are undoubtedly more useful, but are they any more conscious?
Skipping over what constitutes consciousness, which is itself a contentious
subject, the general consensus is no. Again there are interesting questions
about what a sentient machine would mean, and while they are philosophically
interesting; I don't see any compelling evidence that modern AI systems are
approaching anything that would be broadly recognised as consciousness.</p>
<h3>Predictions</h3>
<ul>
<li><strong>1 Year</strong>: Focus shifts towards real world problems caused by the adoption
of AI instead of longer term existential threats. Coverage of
AI-related misinformation, fraud, and labour disputes will exceed
coverage of existential AI risk in mainstream media.</li>
<li><strong>2 Years</strong>: People will increasingly personify chatbots leading to an
increase in social issues related to AI, however this will
likely be over reported relative to the real life scale of the
problem.</li>
<li><strong>5 Years</strong>: Artificial General Intelligence remains hypothetical, with no
broadly accepted examples emerging over the next five years.</li>
</ul>
<h2>Corporations and centralisation</h2>
<p>One trend that has been clear over the past few years is an increase in <a href="https://en.wikipedia.org/wiki/As_a_service">X as a
service</a> business models. With this has come an increasing
reliance on centralised services and consumer subscription models; AI
is no exception to this trend. A common phrase from ~10 years ago was:</p>
<blockquote>
<p>The Cloud Is Just Someone Else’s Computer</p>
</blockquote>
<p>While fundamentally true, for workloads related to AI models most people do not
have hardware capable of running (or training) them. It is certainly possible
to run local models with projects like <a href="https://ollama.com/">Ollama</a>, however consumer
hardware with sufficient RAM is frequently a limiting factor. This is only
exacerbated by demand outstripping supply as new data centres are built.</p>
<p>The reality for most users today is AI models are only accessible remotely
through services like ChatGPT, Claude, and Gemini. This brings with it a number
of concerns around privacy, control, and censorship. There are some niche
services like <a href="https://confer.to/">Confer</a> aiming to address these problems, but for most
users relying on large vendor subscriptions for access to AI models will be an
inevitability.</p>
<h3>Predictions</h3>
<ul>
<li><strong>1 Year</strong>: Consumer hardware prices will not fall, keeping self-hosted AI
out of reach for many people.</li>
<li><strong>2 Years</strong>: Large service providers like ChatGPT and Gemini will increase the
price of subscription models and/or reduce the value of free
tiers by limiting access or increasing ad support.</li>
<li><strong>5 Years</strong>: One of the major AI services (ChatGPT, Claude, or Gemini) will
lose significant market share in response to changes in
subscription models.</li>
</ul>
<h2>Source Code and Security</h2>
<blockquote>
<p>"Given enough eyeballs, all bugs are shallow."</p>
</blockquote>
<p>Software development has already been impacted significantly by the widespread
adoption of LLMs, thanks in part to the widespread availability of free and
open-source code used as training data. This has opened up a number of
interesting questions around copyright, security, and open-source software
development.</p>
<p>Copyright has been a contentious topic for AI, <a href="https://en.wikipedia.org/wiki/The_Pile_(dataset)">significant volumes of
copyrighted material</a> are frequently used in training
data sets and in some instances <a href="https://arxiv.org/abs/2505.12546">can be fully reproduced</a>.
However for software development, big companies like Microsoft have put their
weight behind LLMs with their <a href="https://blogs.microsoft.com/on-the-issues/2023/09/07/copilot-copyright-commitment-ai-legal-concerns/">Copilot Copyright
Commitment</a>. Consequently the threat of legal
action isn't a big concern for many companies adopting LLMs to generate code.</p>
<p>The biggest driving factor behind the adoption of LLMs in software development
will ultimately be efficiency. <a href="https://metr.org/">Metr</a> have done some
<a href="https://arxiv.org/pdf/2507.09089">interesting research</a> looking at this. Counter intuitively
the study found open-source developers were actually more efficient when not
using AI despite perceiving the opposite. They have since done <a href="https://metr.org/blog/2026-02-24-uplift-update/#wider-adoption-of-ai-has-made-it-more-difficult-to-measure-task-level-productivity">followup
research</a> showing a small reversal in this trend, but there
is still a lot of uncertainty around how much efficiency is gained using LLMs.
Despite this, many companies are looking to reduce their developer workforce,
recently Atlassian laid off <a href="https://www.atlassian.com/blog/company-news/atlassian-team-update-march-2026">~1600 employees</a> stating:</p>
<blockquote>
<p>Our approach is not “AI replaces people”. But it would be disingenuous to
pretend AI doesn’t change the mix of skills we need or the number of roles
required in certain areas. It does.</p>
</blockquote>
<p>Not to be left out, LLMs are also beginning to have a big impact on security
research. With the recent <a href="https://red.anthropic.com/2026/mythos-preview/">announcement of Claude Mythos</a>,
there has been a lot of speculation; fueled by statements like "we view this as
a watershed moment for security", and limited access to the model to "critical
industry partners and open-source developers". Unsurprisingly this has
generated headline news, which in many cases is overly alarmist. With that in
mind, I've found Daniel Stenberg's recent post "<a href="https://daniel.haxx.se/blog/2026/05/11/mythos-finds-a-curl-vulnerability/">Mythos finds a curl
vulnerability</a>" to be a much more level headed take on where
using AI models to find vulnerabilities currently is.</p>
<h3>Predictions</h3>
<ul>
<li><strong>1 Year</strong>: Companies aiming to reduce their headcount by replacing
developers with AI will run into troubled waters. To give a
concrete example I suspect Atlassian will either look to rehire
roles in software research and development, or face falling stock
prices.</li>
<li><strong>2 Years</strong>: 2027 will see a marked increase in <a href="https://www.cve.org/about/Metrics">published
CVEs</a>, partially driven by the adoption of AI in
security research. This will go hand in hand with the
strengthening of security, particularly in open-source projects.
Unfortunately this will likely also be coupled with an increase
in low quality bug reports, similar to reports which led to curl
<a href="https://daniel.haxx.se/blog/2026/01/26/the-end-of-the-curl-bug-bounty/">closing their bug-bounty program</a>.</li>
<li><strong>5 Years</strong>: While LLMs are used routinely in software development and
security research, people still play a key role. It is still not
possible to develop and maintain complex mature code bases, or
discover and remediate security vulnerabilities without detailed
technical knowledge.</li>
</ul>
<h2>Time Will Tell</h2>
<p>So how confident am I in my predictions? Not very...</p>
<p>Predicting the future is a notoriously difficult task, and I'm sure I'll look
back and laugh at some of the predictions above. With that said, the one thing
I am confident about is watching the next few years unfold is going to be
interesting.</p>
</article>
<footer class="footer pure-u-1 pure-u-md-3-4">
<hr />
<small>
Copyright © 2026 Floating Octothorpe. Except where
otherwise noted, content on this site is licensed under a
<a rel="license"
href="https://creativecommons.org/licenses/by/4.0/">Creative Commons
Attribution 4.0 License</a>.
</small>
</footer>
</body>
</html>