AI hype isn’t helping anyone
We've got real issues with AI, and security is top of the list. Prompt injecting allows hackers to trick your private LLM, and that's a serious problem.
Stories by Matt Asay
-
-
Reining in the BS in AI
Large language models trained on questionable stuff online will produce more of the same. Retrieval augmented generation is one way to get closer to truth.
-
GitHub’s all-in bet on AI may overlook Git
Not everyone wants AI to do everything for them. Will the risk of losing transparency and visibility into code change how GitHub made collaborative coding so powerful?
-
KubeCon points to the future of enterprise IT
eBPF and Cilium mature into de facto cloud infrastructure, Tetragon tackles new security domains, while OTel topples walled gardens in telemetry data.
-
The clouds can’t afford to forget developers
In the rush to AI, vendors should remember that developers have a lot of clout with IT spending. Helping developers will translate to growth at the bottom line.
-
What AI won’t replace in your programming
Generative AI is great at handling tedium and finding errors, but the expertise and intuition of programmers will always be essential.
-
Making sure open source doesn’t fail AI
The lessons learned from cloud are spurring a proactive examination of what it means to be 'open source' in the rapidly evolving world of AI.
-
Linux distros need to take more responsibility for security
Between the rapid release of open source software, and modern OSes preloaded with packages, enterprises are vulnerable to attacks they aren’t even aware of.
-
People and Python in AI
If you want to squeeze the most value from your data, teach your employees Python and Excel instead of specialised programming languages.
-
Testing the limits of generative AI
As part of the learning curve with AI and LLMs, experiment all you want, but take the results with some skepticism, especially if you’re using it to write your code.
-
A new way of thinking about open source sustainability
Go beyond paying developers to maintain the software your business depends on. Pay the companies that pay the developers and watch the whole ecosystem thrive.
-
Rethinking open source for AI
What are the license rights and restrictions for large language models? Do they cover weights and deep neural network architectures?
-
When the generative AI hype fades
GenAI is a small piece of the artificial intelligence pie, not the whole pie itself. Keep paying attention to deep learning and machine learning.
-
Follow the cloud money
Led by CIQ, Oracle, and SUSE, the new OpenELA trade association is likely to fail without at least one major cloud vendor on board.
-
HashiCorp’s software license turns realpolitik
The move to the Business Source License will allow HashiCorp to continue investing in its products and will force big vendors to become better partners, a win for all users.
