Some interesting links that I Tweeted about in the last week (I also post these on Mastodon, Threads, Newsmast, Bluesky and Post):
- Anyone with even the slightest knowledge of the technology being used could have told them that 1) It was going to have bad performance over women of colour, and 2) That the people using it would be too lazy to double-check: https://www.nzherald.co.nz/nz/supermarket-facial-recognition-trial-rotorua-mothers-discrimination-ordeal/IK4ZEJHLQVFRLMDE6LX4AR57PE/
- An optical neural network chip uses 0.1% of the energy of other chips: https://spectrum.ieee.org/optical-neural-network
- One trial finds five hours per week of productivity savings from using AI: https://www.newshub.co.nz/home/money/2024/04/artificial-intelligence-trial-shows-ai-can-result-in-up-to-five-hours-a-week-of-beautiful-productivity-savings.html
- As I predicted, this AI powered badge is nowhere near as useful as its makers claimed it would be: https://www.theverge.com/24126502/humane-ai-pin-review
- Jailbreaking ChatGPT, and how it helps make AI safer: https://www.freethink.com/robots-ai/chatgpt-jailbreakers
- Are self-replicating AI just around the corner? I don't think so, their power demands are too large at the moment: https://futurism.com/the-byte/anthropic-ceo-ai-replicate-survive
- An AI doorcam that shoots paintballs. This just looks like a lawsuit waiting to happen: https://newatlas.com/technology/paintball-security-paintcam-eve/
- I don't think making AI chatbots of deceased loved ones is a good idea. The chatbots will never be true to the person, and they could stop people moving on with their lives after a loss: https://www.freethink.com/robots-ai/thanabots
- Meta's AI claims to have a disabled and gifted child: https://futurism.com/the-byte/meta-ai-child
- The best CEOs are people people, who can really lead and motivate their staff. I don't see AI being able to effectively do that, especially if people know they're AI: https://futurism.com/the-byte/ceos-fear-ai-jobs
- A graphical representation of 15 statistics about AI: https://spectrum.ieee.org/ai-index-2024
- An interesting article on how machine learning was used to learn the mechanics of insect wing hinges: https://www.nature.com/articles/s41586-024-07293-4
- On the need for new metrics to measure the performance of AI: https://www.nature.com/articles/d41586-024-01087-4
- Hala Point, a neuromorphic AI architecture from Intel: https://www.datanami.com/2024/04/18/intel-announces-hala-point-worlds-largest-neuromorphic-system-for-sustainable-ai/
- Meta releases their latest LLM AI model, claims it's the best one yet: https://www.theverge.com/2024/4/18/24133808/meta-ai-assistant-llama-3-chatgpt-openai-rival
- An AI heavy automated fast food place is underwhelming: https://www.theguardian.com/technology/2024/apr/15/ai-burger-joint-flippy-caliexpress I do expect these to get better, and more common, in the future
- Using machine learning to detect small differences in brain activity indicative of psychosis: https://www.extremetech.com/science/machine-learning-helps-scientists-locate-the-neurological-origin-of-psychosis
- Content generated by AI is NOT a passing fad: https://www.theregister.com/2024/04/17/devaluing_ai_content_is_lazy/?td=rt-4a
- Educators' unions are having regulations around generative AI written into contracts: https://www.insidehighered.com/news/tech-innovation/artificial-intelligence/2024/04/18/faculty-unions-seek-ai-guidelines-contracts
- How AI is affecting retail: https://www.informationweek.com/machine-learning-ai/how-ai-is-reshaping-retail
- More expansion in data centres to support growth of AI: https://www.theregister.com/2024/04/18/microsoft_datacenter_capacity/ Where is the energy to run and cool these centres going to come from?
- No, Meta's labelling of AI generated content will not stop AI being used to produce and spread misinformation: https://www.extremetech.com/internet/meta-will-begin-labeling-ai-generated-content-misinformation
- AI is good for small businesses as well: https://dataconomy.com/2024/04/18/pwc-global-nextgen-survey-2024/
- AI is definitely in a bubble, the question is, how much of a mess will it make when it bursts? https://www.kdnuggets.com/dot-com-to-dot-ai-new-tech-bubble
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.